9129767 EY7L9SEU 1 apa 50 date desc year Gerstoft 18 https://pgerstoft.scrippsprofiles.ucsd.edu/wp-content/plugins/zotpress/
%7B%22status%22%3A%22success%22%2C%22updateneeded%22%3Afalse%2C%22instance%22%3Afalse%2C%22meta%22%3A%7B%22request_last%22%3A200%2C%22request_next%22%3A50%2C%22used_cache%22%3Atrue%7D%2C%22data%22%3A%5B%7B%22key%22%3A%22LH6SYXB4%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Jenkins%20et%20al.%22%2C%22parsedDate%22%3A%222024-08-01%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EJenkins%2C%20W.%20F.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20%26amp%3B%20Park%2C%20Y.%20%282024%29.%20Geoacoustic%20inversion%20using%20Bayesian%20optimization%20with%20a%20Gaussian%20process%20surrogate%20model.%20%3Ci%3EThe%20Journal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E156%3C%5C%2Fi%3E%282%29%2C%20812%26%23x2013%3B822.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0028177%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0028177%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Geoacoustic%20inversion%20using%20Bayesian%20optimization%20with%20a%20Gaussian%20process%20surrogate%20model%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22William%20F.%22%2C%22lastName%22%3A%22Jenkins%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Yongsung%22%2C%22lastName%22%3A%22Park%22%7D%5D%2C%22abstractNote%22%3A%22Geoacoustic%20inversion%20can%20be%20a%20computationally%20expensive%20task%20in%20high-dimensional%20parameter%20spaces%2C%20typically%20requiring%20thousands%20of%20forward%20model%20evaluations%20to%20estimate%20the%20geoacoustic%20environment.%20We%20demonstrate%20Bayesian%20optimization%20%28BO%29%2C%20an%20efficient%20global%20optimization%20method%20capable%20of%20estimating%20geoacoustic%20parameters%20in%20seven-dimensional%20space%20within%20100%20evaluations%20instead%20of%20thousands.%20BO%20iteratively%20searches%20parameter%20space%20for%20the%20global%20optimum%20of%20an%20objective%20function%2C%20defined%20in%20this%20study%20as%20the%20Bartlett%20power.%20Each%20step%20consists%20of%20fitting%20a%20Gaussian%20process%20surrogate%20model%20to%20observed%20data%20and%20then%20choosing%20a%20new%20point%20to%20evaluate%20using%20a%20heuristic%20acquisition%20function.%20The%20ideal%20acquisition%20function%20balances%20exploration%20of%20the%20parameter%20space%20in%20regions%20with%20high%20uncertainty%20with%20exploitation%20of%20high-performing%20regions.%20Three%20acquisition%20functions%20are%20evaluated%3A%20upper%20confidence%20bound%2C%20expected%20improvement%20%28EI%29%2C%20and%20logarithmically%20transformed%20EI.%20BO%20is%20demonstrated%20for%20both%20simulated%20and%20experimental%20data%20from%20a%20shallow-water%20environment%20and%20rapidly%20estimates%20optimal%20parameters%20while%20yielding%20results%20comparable%20to%20differential%20evolution%20optimization.%22%2C%22date%22%3A%222024-08-01%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0028177%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fpubs.aip.org%5C%2Fjasa%5C%2Farticle%5C%2F156%5C%2F2%5C%2F812%5C%2F3307198%5C%2FGeoacoustic-inversion-using-Bayesian-optimization%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%22A6U9IMQY%22%2C%2289L3J9FA%22%5D%2C%22dateModified%22%3A%222024-08-30T21%3A16%3A17Z%22%7D%7D%2C%7B%22key%22%3A%22WIPJ3J3R%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Zhou%20et%20al.%22%2C%22parsedDate%22%3A%222024-07-08%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EZhou%2C%20Z.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20%26amp%3B%20Olsen%2C%20K.%20%282024%29.%20Graph-learning%20approach%20to%20combine%20multiresolution%20seismic%20velocity%20models.%20%3Ci%3EGeophysical%20Journal%20International%3C%5C%2Fi%3E%2C%20%3Ci%3E238%3C%5C%2Fi%3E%283%29%2C%201353%26%23x2013%3B1365.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1093%5C%2Fgji%5C%2Fggae212%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1093%5C%2Fgji%5C%2Fggae212%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Graph-learning%20approach%20to%20combine%20multiresolution%20seismic%20velocity%20models%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Zheng%22%2C%22lastName%22%3A%22Zhou%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Kim%22%2C%22lastName%22%3A%22Olsen%22%7D%5D%2C%22abstractNote%22%3A%22SUMMARY%5Cn%20%20%20%20%20%20%20%20%20%20%20%20The%20resolution%20of%20velocity%20models%20obtained%20by%20tomography%20varies%20due%20to%20multiple%20factors%20and%20variables%2C%20such%20as%20the%20inversion%20approach%2C%20ray%20coverage%2C%20data%20quality%2C%20etc.%20Combining%20velocity%20models%20with%20different%20resolutions%20can%20enable%20more%20accurate%20ground%20motion%20simulations.%20Toward%20this%20goal%2C%20we%20present%20a%20novel%20methodology%20to%20fuse%20multiresolution%20seismic%20velocity%20maps%20with%20probabilistic%20graphical%20models%20%28PGMs%29.%20The%20PGMs%20provide%20segmentation%20results%2C%20corresponding%20to%20various%20velocity%20intervals%2C%20in%20seismic%20velocity%20models%20with%20different%20resolutions.%20Further%2C%20by%20considering%20physical%20information%20%28such%20as%20ray%20path%20density%29%2C%20we%20introduce%20physics-informed%20probabilistic%20graphical%20models%20%28PIPGMs%29.%20These%20models%20provide%20data-driven%20relations%20between%20subdomains%20with%20low%20%28LR%29%20and%20high%20%28HR%29%20resolutions.%20Transferring%20%28segmented%29%20distribution%20information%20from%20the%20HR%20regions%20enhances%20the%20details%20in%20the%20LR%20regions%20by%20solving%20a%20maximum%20likelihood%20problem%20with%20prior%20knowledge%20from%20HR%20models.%20When%20updating%20areas%20bordering%20HR%20and%20LR%20regions%2C%20a%20patch-scanning%20policy%20is%20adopted%20to%20consider%20local%20patterns%20and%20avoid%20sharp%20boundaries.%20To%20evaluate%20the%20efficacy%20of%20the%20proposed%20PGM%20fusion%20method%2C%20we%20tested%20the%20fusion%20approach%20on%20both%20a%20synthetic%20checkerboard%20model%20and%20a%20fault%20zone%20structure%20imaged%20from%20the%202019%20Ridgecrest%2C%20CA%2C%20earthquake%20sequence.%20The%20Ridgecrest%20fault%20zone%20image%20consists%20of%20a%20shallow%20%28top%201%5Cu00a0km%29%20high-resolution%20shear-wave%20velocity%20model%20obtained%20from%20ambient%20noise%20tomography%2C%20which%20is%20embedded%20into%20the%20coarser%20Statewide%20California%20Earthquake%20Center%20Community%20Velocity%20Model%20version%20S4.26-M01.%20The%20model%20efficacy%20is%20underscored%20by%20the%20deviation%20between%20observed%20and%20calculated%20traveltimes%20along%20the%20boundaries%20between%20HR%20and%20LR%20regions%2C%2038%5Cu00a0per%5Cu00a0cent%20less%20than%20obtained%20by%20conventional%20Gaussian%20interpolation.%20The%20proposed%20PGM%20fusion%20method%20can%20merge%20any%20gridded%20multiresolution%20velocity%20model%2C%20a%20valuable%20tool%20for%20computational%20seismology%20and%20ground%20motion%20estimation.%22%2C%22date%22%3A%222024-07-08%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1093%5C%2Fgji%5C%2Fggae212%22%2C%22ISSN%22%3A%220956-540X%2C%201365-246X%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Facademic.oup.com%5C%2Fgji%5C%2Farticle%5C%2F238%5C%2F3%5C%2F1353%5C%2F7693745%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222024-07-30T21%3A18%3A50Z%22%7D%7D%2C%7B%22key%22%3A%22ACH4CNX8%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Li%20et%20al.%22%2C%22parsedDate%22%3A%222024-07-01%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3ELi%2C%20J.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20%26amp%3B%20Fan%2C%20J.%20%282024%29.%20Eigenvalues%20of%20the%20noise%20covariance%20matrix%20in%20ocean%20waveguides.%20%3Ci%3EThe%20Journal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E156%3C%5C%2Fi%3E%281%29%2C%20189%26%23x2013%3B201.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0026477%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0026477%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Eigenvalues%20of%20the%20noise%20covariance%20matrix%20in%20ocean%20waveguides%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Jie%22%2C%22lastName%22%3A%22Li%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Jun%22%2C%22lastName%22%3A%22Fan%22%7D%5D%2C%22abstractNote%22%3A%22The%20eigenvalue%20%28EV%29%20spectra%20of%20the%20theoretical%20noise%20covariance%20matrix%20%28CM%29%20and%20observed%20sample%20CM%20provide%20information%20about%20the%20environment%2C%20source%2C%20and%20noise%20generation.%20This%20paper%20investigates%20these%20spectra%20for%20vertical%20line%20arrays%20%28VLAs%29%20and%20horizontal%20line%20arrays%20%28HLAs%29%20in%20deep%20and%20shallow%20water%20numerically.%20Empirically%2C%20the%20spectra%20are%20related%20to%20the%20width%20of%20the%20conventional%20beamforming%20output%20in%20angle%20space.%20In%20deep%20water%2C%20the%20HLA%20noise%20CM%20tends%20to%20be%20isotropic%20regardless%20of%20the%20sound%20speed%20profile.%20Thus%2C%20the%20EV%20spectrum%20approaches%20a%20step%20function.%20In%20contrast%2C%20the%20VLA%20noise%20CM%20is%20non-isotropic%2C%20and%20the%20EVs%20of%20the%20CM%20jump%20in%20two%20steps.%20The%20EVs%20before%20the%20first%20jump%20are%20due%20to%20sea%20surface%20noise%2C%20while%20those%20between%20the%20first%20and%20second%20jump%20are%20due%20to%20bottom-reflected%20noise.%20In%20shallow%20water%2C%20the%20VLA%20noise%20CM%20is%20affected%20by%20the%20environment%20%28sound%20speed%20profile%20and%20seabed%20density%2C%20sound%20speed%2C%20attenuation%2C%20and%20layers%29%20and%20array%20depth%2C%20the%20EVs%20have%20a%20more%20complicated%20structure.%20For%20Noise09%20VLA%20experimental%20data%2C%20the%20noise%20sample%20CM%20EVs%20match%20the%20waveguide%20noise%20model%20better%20than%20the%20three-dimensional%20isotropic%20noise%20model.%22%2C%22date%22%3A%222024-07-01%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0026477%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fpubs.aip.org%5C%2Fjasa%5C%2Farticle%5C%2F156%5C%2F1%5C%2F189%5C%2F3302405%5C%2FEigenvalues-of-the-noise-covariance-matrix-in%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222024-07-30T21%3A15%3A28Z%22%7D%7D%2C%7B%22key%22%3A%228NP8HCRI%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Liu%20and%20Gerstoft%22%2C%22parsedDate%22%3A%222024-06-01%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3ELiu%2C%20R.%2C%20%26amp%3B%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%20%282024%29.%20Spatial%20acoustic%20properties%20recovery%20with%20deep%20learning.%20%3Ci%3EThe%20Journal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E155%3C%5C%2Fi%3E%286%29%2C%203690%26%23x2013%3B3701.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0026231%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0026231%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Spatial%20acoustic%20properties%20recovery%20with%20deep%20learning%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Ruixian%22%2C%22lastName%22%3A%22Liu%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%5D%2C%22abstractNote%22%3A%22The%20physics-informed%20neural%20network%20%28PINN%29%20can%20recover%20partial%20differential%20equation%20%28PDE%29%20coefficients%20that%20remain%20constant%20throughout%20the%20spatial%20domain%20directly%20from%20measurements.%20We%20propose%20a%20spatially%20dependent%20physics-informed%20neural%20network%20%28SD-PINN%29%2C%20which%20enables%20recovering%20coefficients%20in%20spatially%20dependent%20PDEs%20using%20one%20neural%20network%2C%20eliminating%20the%20requirement%20for%20domain-specific%20physical%20expertise.%20The%20network%20is%20trained%20by%20minimizing%20a%20combination%20of%20loss%20functions%20involving%20data-fitting%20and%20physical%20constraints%2C%20in%20which%20the%20requirement%20for%20satisfying%20the%20assumed%20governing%20PDE%20is%20encoded.%20For%20the%20recovery%20of%20spatially%20two-dimensional%20%282D%29%20PDEs%2C%20we%20store%20the%20PDE%20coefficients%20at%20all%20locations%20in%20the%202D%20region%20of%20interest%20into%20a%20matrix%20and%20incorporate%20a%20low-rank%20assumption%20for%20this%20matrix%20to%20recover%20the%20coefficients%20at%20locations%20without%20measurements.%20We%20apply%20the%20SD-PINN%20to%20recovering%20spatially%20dependent%20coefficients%20of%20the%20wave%20equation%20to%20reveal%20the%20spatial%20distribution%20of%20acoustic%20properties%20in%20the%20inhomogeneous%20medium.%22%2C%22date%22%3A%222024-06-01%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0026231%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fpubs.aip.org%5C%2Fjasa%5C%2Farticle%5C%2F155%5C%2F6%5C%2F3690%5C%2F3296939%5C%2FSpatial-acoustic-properties-recovery-with-deep%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222024-10-14T22%3A08%3A42Z%22%7D%7D%2C%7B%22key%22%3A%22K6CT4TIW%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Zhou%20et%20al.%22%2C%22parsedDate%22%3A%222024-06-01%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EZhou%2C%20Z.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20%26amp%3B%20Olsen%2C%20K.%20B.%20%282024%29.%203D%20Multiresolution%20Velocity%20Model%20Fusion%20with%20Probability%20Graphical%20Models.%20%3Ci%3EBulletin%20of%20the%20Seismological%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E114%3C%5C%2Fi%3E%283%29%2C%201279%26%23x2013%3B1292.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1785%5C%2F0120230271%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1785%5C%2F0120230271%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%223D%20Multiresolution%20Velocity%20Model%20Fusion%20with%20Probability%20Graphical%20Models%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Zheng%22%2C%22lastName%22%3A%22Zhou%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Kim%20Bak%22%2C%22lastName%22%3A%22Olsen%22%7D%5D%2C%22abstractNote%22%3A%22ABSTRACT%5Cn%20%20%20%20%20%20%20%20%20%20%20%20The%20variability%20in%20spatial%20resolution%20of%20seismic%20velocity%20models%20obtained%20via%20tomographic%20methodologies%20is%20attributed%20to%20many%20factors%2C%20including%20inversion%20strategies%2C%20ray-path%20coverage%2C%20and%20data%20integrity.%20Integration%20of%20such%20models%2C%20with%20distinct%20resolutions%2C%20is%20crucial%20during%20the%20refinement%20of%20community%20models%2C%20thereby%20enhancing%20the%20precision%20of%20ground-motion%20simulations.%20Toward%20this%20goal%2C%20we%20introduce%20the%20probability%20graphical%20model%20%28PGM%29%2C%20combining%20velocity%20models%20with%20heterogeneous%20resolutions%20and%20nonuniform%20data%20point%20distributions.%20The%20PGM%20integrates%20data%20relations%20across%20varying%20resolution%20subdomains%2C%20enhancing%20detail%20within%20low-resolution%20%28LR%29%20domains%20by%20utilizing%20information%20and%20prior%20knowledge%20from%20high-resolution%20%28HR%29%20subdomains%20through%20a%20maximum%20posterior%20problem.%20Assessment%20of%20efficacy%2C%20utilizing%20both%202D%20and%203D%20velocity%20models%5Cu2014consisting%20of%20synthetic%20checkerboard%20models%20and%20a%20fault-zone%20model%20from%20Ridgecrest%2C%20California%5Cu2014demonstrates%20noteworthy%20improvements%20in%20accuracy%2C%20compared%20to%20state-of-the-art%20fusion%20techniques.%20Specifically%2C%20we%20find%20reductions%20of%2030%25%20and%2044%25%20in%20computed%20travel-time%20residuals%20for%202D%20and%203D%20models%2C%20respectively%2C%20as%20compared%20to%20conventional%20smoothing%20techniques.%20Unlike%20conventional%20methods%2C%20the%20PGM%5Cu2019s%20adaptive%20weight%20selection%20facilitates%20preserving%20and%20learning%20details%20from%20complex%2C%20nonuniform%20HR%20models%20and%20applies%20the%20enhancements%20to%20the%20LR%20background%20domain.%22%2C%22date%22%3A%222024-06-01%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1785%5C%2F0120230271%22%2C%22ISSN%22%3A%220037-1106%2C%201943-3573%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fpubs.geoscienceworld.org%5C%2Fbssa%5C%2Farticle%5C%2F114%5C%2F3%5C%2F1279%5C%2F638808%5C%2F3D-Multiresolution-Velocity-Model-Fusion-with%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222024-07-30T18%3A18%3A18Z%22%7D%7D%2C%7B%22key%22%3A%22P96AN969%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Wiens%20et%20al.%22%2C%22parsedDate%22%3A%222024-04-16%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EWiens%2C%20D.%20A.%2C%20Aster%2C%20R.%20C.%2C%20Nyblade%2C%20A.%20A.%2C%20Bromirski%2C%20P.%20D.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20%26amp%3B%20Stephen%2C%20R.%20A.%20%282024%29.%20Ross%20Ice%20Shelf%20Displacement%20and%20Elastic%20Plate%20Waves%20Induced%20by%20Whillans%20Ice%20Stream%20Slip%20Events.%20%3Ci%3EGeophysical%20Research%20Letters%3C%5C%2Fi%3E%2C%20%3Ci%3E51%3C%5C%2Fi%3E%287%29%2C%20e2023GL108040.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1029%5C%2F2023GL108040%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1029%5C%2F2023GL108040%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Ross%20Ice%20Shelf%20Displacement%20and%20Elastic%20Plate%20Waves%20Induced%20by%20Whillans%20Ice%20Stream%20Slip%20Events%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Douglas%20A.%22%2C%22lastName%22%3A%22Wiens%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Richard%20C.%22%2C%22lastName%22%3A%22Aster%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Andrew%20A.%22%2C%22lastName%22%3A%22Nyblade%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%20D.%22%2C%22lastName%22%3A%22Bromirski%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Ralph%20A.%22%2C%22lastName%22%3A%22Stephen%22%7D%5D%2C%22abstractNote%22%3A%22Abstract%5Cn%20%20%20%20%20%20%20%20%20%20%20%20%5Cn%20%20%20%20%20%20%20%20%20%20%20%20%20%20Ice%20shelves%20are%20assumed%20to%20flow%20steadily%20from%20their%20grounding%20lines%20to%20the%20ice%20front.%20We%20report%20the%20detection%20of%20ice%5Cu2010propagating%20extensional%20Lamb%20%28plate%29%20waves%20accompanied%20by%20pulses%20of%20permanent%20ice%20shelf%20displacement%20observed%20by%20co%5Cu2010located%20Global%20Navigation%20Satellite%20System%5Cu00a0receivers%20and%20seismographs%20on%20the%20Ross%20Ice%20Shelf.%20The%20extensional%20waves%20and%20associated%20ice%20shelf%20displacement%20are%20produced%20by%20tidally%20triggered%20basal%20slip%20events%20of%20the%20Whillans%20Ice%20Stream%2C%20which%20flows%20into%20the%20ice%20shelf.%20The%20propagation%20velocity%20of%202%2C800%5Cu00a0m%5C%2Fs%20is%20intermediate%20between%20shear%20and%20compressional%20ice%20velocities%2C%20with%20velocity%20and%20particle%20motions%20consistent%20with%20predictions%20for%20extensional%20Lamb%20waves.%20During%20the%20passage%20of%20the%20Lamb%20waves%20the%20entire%20ice%20shelf%20is%20displaced%20about%2060%5Cu00a0mm%20with%20a%20velocity%20more%20than%20an%20order%20of%20magnitude%20above%20its%20long%5Cu2010term%20flow%20rate.%20Observed%20displacements%20indicate%20a%20peak%20dynamic%20strain%20of%2010%5Cn%20%20%20%20%20%20%20%20%20%20%20%20%20%20%5Cu22127%5Cn%20%20%20%20%20%20%20%20%20%20%20%20%20%20%2C%20comparable%20to%20that%20of%20earthquake%20surface%20waves%20that%20trigger%20ice%20quakes.%5Cn%20%20%20%20%20%20%20%20%20%20%20%20%5Cn%20%20%20%20%20%20%20%20%20%20%2C%20%5Cn%20%20%20%20%20%20%20%20%20%20%20%20Plain%20Language%20Summary%5Cn%20%20%20%20%20%20%20%20%20%20%20%20Ice%20shelves%20normally%20flow%20steadily%20toward%20their%20boundaries%20with%20the%20open%20ocean%20at%20the%20ice%20front.%20However%2C%20seismographs%20and%20Global%20Navigation%20Satellite%20System%20receivers%20deployed%20on%20the%20Ross%20Ice%20Shelf%20record%20guided%20elastic%20plate%20waves%20traveling%20in%20the%20ice%20as%20well%20as%20permanent%20displacement%20of%20the%20ice%20shelf.%20The%20elastic%20waves%20and%20ice%20shelf%20displacement%20originate%20from%20basal%20slip%20events%20of%20the%20Whillans%20Ice%20Stream%2C%20which%20flows%20into%20the%20Ross%20Ice%20Shelf.%20The%20velocity%20of%20the%20elastic%20waves%20is%20about%202%2C800%5Cu00a0m%5C%2Fs%2C%20as%20expected%20for%20guided%20plate%20waves%20propagating%20in%20an%20ice%20shelf.%20During%20the%20passage%20of%20the%20elastic%20waves%2C%20the%20entire%20ice%20shelf%20with%20an%20area%20of%20500%2C000%20square%20kilometers%20is%20displaced%20about%2060%5Cu00a0mm%20in%20a%20direction%20away%20from%20the%20Whillans%20Ice%20Stream.%20These%20observations%20show%20that%20the%20strain%20imparted%20to%20the%20ice%20shelf%20by%20the%20once%20or%20twice%20daily%20Whillans%20Ice%20Stream%20basal%20slip%20events%20is%20sufficient%20to%20trigger%20ice%20quakes%20and%20perhaps%20enhance%20the%20deformation%20of%20the%20ice%20shelf.%5Cn%20%20%20%20%20%20%20%20%20%20%2C%20%5Cn%20%20%20%20%20%20%20%20%20%20%20%20Key%20Points%5Cn%20%20%20%20%20%20%20%20%20%20%20%20%5Cn%20%20%20%20%20%20%20%20%20%20%20%20%20%20%5Cn%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%5Cn%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20Extensional%20Lamb%20waves%20propagate%20across%20the%20Ross%20Ice%20Shelf%2C%20radiated%20from%20slip%20events%20at%20the%20base%20of%20the%20Whillans%20Ice%20Stream%5Cn%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%5Cn%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%5Cn%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20During%20the%20passage%20of%20the%20Lamb%20waves%2C%20the%20entire%20ice%20shelf%20is%20displaced%20about%2060%5Cu00a0mm%2C%20with%20a%20velocity%20an%20order%20of%20magnitude%20above%20its%20long%5Cu2010term%20flow%20rate%5Cn%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%5Cn%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%5Cn%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%5Cn%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20The%20displacement%20pulses%20produce%20a%20peak%20dynamic%20strain%20of%2010%5Cn%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%5Cu22127%5Cn%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%20%2C%20suggesting%20that%20they%20could%20trigger%20icequakes%20in%20the%20ice%20shelf%22%2C%22date%22%3A%222024-04-16%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1029%5C%2F2023GL108040%22%2C%22ISSN%22%3A%220094-8276%2C%201944-8007%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fagupubs.onlinelibrary.wiley.com%5C%2Fdoi%5C%2F10.1029%5C%2F2023GL108040%22%2C%22collections%22%3A%5B%22D2UAFEX5%22%2C%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222024-04-12T20%3A54%3A06Z%22%7D%7D%2C%7B%22key%22%3A%22GJW99MNC%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Yoon%20et%20al.%22%2C%22parsedDate%22%3A%222024-03-01%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EYoon%2C%20S.%2C%20Park%2C%20Y.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20%26amp%3B%20Seong%2C%20W.%20%282024%29.%20Predicting%20ocean%20pressure%20field%20with%20a%20physics-informed%20neural%20network.%20%3Ci%3EThe%20Journal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E155%3C%5C%2Fi%3E%283%29%2C%202037%26%23x2013%3B2049.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0025235%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0025235%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Predicting%20ocean%20pressure%20field%20with%20a%20physics-informed%20neural%20network%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Seunghyun%22%2C%22lastName%22%3A%22Yoon%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Yongsung%22%2C%22lastName%22%3A%22Park%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Woojae%22%2C%22lastName%22%3A%22Seong%22%7D%5D%2C%22abstractNote%22%3A%22Ocean%20sound%20pressure%20field%20prediction%2C%20based%20on%20partially%20measured%20pressure%20magnitudes%20at%20different%20range-depths%2C%20is%20presented.%20Our%20proposed%20machine%20learning%20strategy%20employs%20a%20trained%20neural%20network%20with%20range-depth%20as%20input%20and%20outputs%20complex%20acoustic%20pressure%20at%20the%20location.%20We%20utilize%20a%20physics-informed%20neural%20network%20%28PINN%29%2C%20fitting%20sampled%20data%20while%20considering%20the%20additional%20information%20provided%20by%20the%20partial%20differential%20equation%20%28PDE%29%20governing%20the%20ocean%20sound%20pressure%20field.%20In%20vast%20ocean%20environments%20with%20kilometer-scale%20ranges%2C%20pressure%20fields%20exhibit%20rapidly%20fluctuating%20phases%2C%20even%20at%20frequencies%20below%20100%5Cu2009Hz%2C%20posing%20a%20challenge%20for%20neural%20networks%20to%20converge%20to%20accurate%20solutions.%20To%20address%20this%2C%20we%20utilize%20the%20envelope%20function%20from%20the%20parabolic-equation%20technique%2C%20fundamental%20in%20ocean%20sound%20propagation%20modeling.%20The%20envelope%20function%20shows%20slower%20variations%20across%20ranges%2C%20enabling%20PINNs%20to%20predict%20sound%20pressure%20in%20an%20ocean%20waveguide%20more%20effectively.%20Additional%20PDE%20information%20allows%20PINNs%20to%20capture%20PDE%20solutions%20even%20with%20a%20limited%20amount%20of%20training%20data%2C%20distinguishing%20them%20from%20purely%20data-driven%20machine%20learning%20approaches%20that%20require%20extensive%20datasets.%20Our%20approach%20is%20validated%20through%20simulations%20and%20using%20data%20from%20the%20SWellEx-96%20experiment.%22%2C%22date%22%3A%222024-03-01%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0025235%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fpubs.aip.org%5C%2Fjasa%5C%2Farticle%5C%2F155%5C%2F3%5C%2F2037%5C%2F3271348%5C%2FPredicting-ocean-pressure-field-with-a-physics%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%22A6U9IMQY%22%5D%2C%22dateModified%22%3A%222024-04-12T20%3A30%3A13Z%22%7D%7D%2C%7B%22key%22%3A%22YS7HRNYR%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Mecklenbr%5Cu00e4uker%20et%20al.%22%2C%22parsedDate%22%3A%222024%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EMecklenbr%26%23xE4%3Buker%2C%20C.%20F.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20Ollila%2C%20E.%2C%20%26amp%3B%20Park%2C%20Y.%20%282024%29.%20Robust%20and%20sparse%20M-estimation%20of%20DOA.%20%3Ci%3ESignal%20Processing%3C%5C%2Fi%3E%2C%20%3Ci%3E220%3C%5C%2Fi%3E%2C%20109461.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1016%5C%2Fj.sigpro.2024.109461%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1016%5C%2Fj.sigpro.2024.109461%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Robust%20and%20sparse%20M-estimation%20of%20DOA%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Christoph%20F.%22%2C%22lastName%22%3A%22Mecklenbr%5Cu00e4uker%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Esa%22%2C%22lastName%22%3A%22Ollila%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Yongsung%22%2C%22lastName%22%3A%22Park%22%7D%5D%2C%22abstractNote%22%3A%22%22%2C%22date%22%3A%2207%5C%2F2024%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1016%5C%2Fj.sigpro.2024.109461%22%2C%22ISSN%22%3A%2201651684%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Flinkinghub.elsevier.com%5C%2Fretrieve%5C%2Fpii%5C%2FS016516842400080X%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%22A6U9IMQY%22%5D%2C%22dateModified%22%3A%222024-04-18T21%3A51%3A35Z%22%7D%7D%2C%7B%22key%22%3A%22DXSUFTPV%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Park%20and%20Gerstoft%22%2C%22parsedDate%22%3A%222024%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EPark%2C%20Y.%2C%20%26amp%3B%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%20%282024%29.%20Difference%20Frequency%20Gridless%20Sparse%20Array%20Processing.%20%3Ci%3EIEEE%20Open%20Journal%20of%20Signal%20Processing%3C%5C%2Fi%3E%2C%20%3Ci%3E5%3C%5C%2Fi%3E%2C%20914%26%23x2013%3B925.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2FOJSP.2024.3425284%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2FOJSP.2024.3425284%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Difference%20Frequency%20Gridless%20Sparse%20Array%20Processing%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Yongsung%22%2C%22lastName%22%3A%22Park%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%5D%2C%22abstractNote%22%3A%22%22%2C%22date%22%3A%222024%22%2C%22language%22%3A%22%22%2C%22DOI%22%3A%2210.1109%5C%2FOJSP.2024.3425284%22%2C%22ISSN%22%3A%222644-1322%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fieeexplore.ieee.org%5C%2Fdocument%5C%2F10591424%5C%2F%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%22A6U9IMQY%22%5D%2C%22dateModified%22%3A%222024-08-29T22%3A48%3A56Z%22%7D%7D%2C%7B%22key%22%3A%22A2PSL2BX%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Wu%20et%20al.%22%2C%22parsedDate%22%3A%222024%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EWu%2C%20Y.%2C%20Wakin%2C%20M.%20B.%2C%20%26amp%3B%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%20%282024%29.%20Non-Uniform%20Array%20and%20Frequency%20Spacing%20for%20Regularization-Free%20Gridless%20DOA.%20%3Ci%3EIEEE%20Transactions%20on%20Signal%20Processing%3C%5C%2Fi%3E%2C%20%3Ci%3E72%3C%5C%2Fi%3E%2C%202006%26%23x2013%3B2020.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2FTSP.2024.3386018%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2FTSP.2024.3386018%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Non-Uniform%20Array%20and%20Frequency%20Spacing%20for%20Regularization-Free%20Gridless%20DOA%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Yifan%22%2C%22lastName%22%3A%22Wu%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Michael%20B.%22%2C%22lastName%22%3A%22Wakin%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%5D%2C%22abstractNote%22%3A%22%22%2C%22date%22%3A%222024%22%2C%22language%22%3A%22%22%2C%22DOI%22%3A%2210.1109%5C%2FTSP.2024.3386018%22%2C%22ISSN%22%3A%221053-587X%2C%201941-0476%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fieeexplore.ieee.org%5C%2Fdocument%5C%2F10494827%5C%2F%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%2289L3J9FA%22%5D%2C%22dateModified%22%3A%222024-07-31T23%3A17%3A50Z%22%7D%7D%2C%7B%22key%22%3A%22TW6DAZ5E%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Sathyanarayanan%20et%20al.%22%2C%22parsedDate%22%3A%222023-11%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3ESathyanarayanan%2C%20V.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20%26amp%3B%20El%20Gamal%2C%20A.%20%282023%29.%20RML22%3A%20Realistic%20Dataset%20Generation%20for%20Wireless%20Modulation%20Classification.%20%3Ci%3EIeee%20Transactions%20on%20Wireless%20Communications%3C%5C%2Fi%3E%2C%20%3Ci%3E22%3C%5C%2Fi%3E%2811%29%2C%207663%26%23x2013%3B7675.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2FTwc.2023.3254490%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2FTwc.2023.3254490%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22RML22%3A%20Realistic%20Dataset%20Generation%20for%20Wireless%20Modulation%20Classification%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22V.%22%2C%22lastName%22%3A%22Sathyanarayanan%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22A.%22%2C%22lastName%22%3A%22El%20Gamal%22%7D%5D%2C%22abstractNote%22%3A%22Application%20of%20Deep%20learning%20%28DL%29%20to%20modulation%20classification%20has%20shown%20significant%20performance%20improvements.%20The%20focus%20has%20been%20model%20centric%2C%20where%20newer%20architectures%20are%20attempted%20on%20benchmark%20dataset%20RADIOML.2016.10A%20%28RML16%29.%20RML16%20is%20a%20high%20impact%20effort%20that%20laid%20the%20foundation%20for%20generating%20a%20synthetic%20dataset%20for%20applying%20DL%20models%20to%20wireless%20problems.%20This%20encouraged%20development%20of%20newer%20architectures%20to%20RML16.%20We%20use%20a%20data%20centric%20DL%20approach%20where%20focus%20moves%20from%20model%20architectures%20to%20data%20quality.%20RML16%20has%20shortcomings%20such%20as%20errors%20and%20ad-hoc%20choices%20of%20parameters.%20We%20build%20upon%20RML16%20and%20provide%20realistic%20and%20correct%20methodology%20of%20generating%20dataset.%20A%20new%20benchmark%20dataset%20RML22%20is%20generated.%20Going%20forward%2C%20we%20envision%20researchers%20to%20improve%20model%20quality%20on%20RML22.%20We%20attempt%20to%20improve%20data%20quality%20by%20studying%20the%20impact%20of%20information%20sources.%20Further%2C%20the%20choices%20of%20artifacts%20and%20signal%20model%20parameterization%20are%20analyzed%20carefully.%20The%20Python%20source%20code%20used%20to%20generate%20RML22%20is%20shared%20to%20enable%20researchers%20to%20further%20improve%20dataset%20quality.%22%2C%22date%22%3A%22Nov%202023%22%2C%22language%22%3A%22English%22%2C%22DOI%22%3A%2210.1109%5C%2FTwc.2023.3254490%22%2C%22ISSN%22%3A%221536-1276%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%2289L3J9FA%22%5D%2C%22dateModified%22%3A%222024-05-03T16%3A55%3A54Z%22%7D%7D%2C%7B%22key%22%3A%22DMP8G6RZ%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Jenkins%20et%20al.%22%2C%22parsedDate%22%3A%222023-09-01%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EJenkins%2C%20W.%20F.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20%26amp%3B%20Park%2C%20Y.%20%282023%29.%20Bayesian%20optimization%20with%20Gaussian%20process%20surrogate%20model%20for%20source%20localization.%20%3Ci%3EThe%20Journal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E154%3C%5C%2Fi%3E%283%29%2C%201459%26%23x2013%3B1470.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0020839%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0020839%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Bayesian%20optimization%20with%20Gaussian%20process%20surrogate%20model%20for%20source%20localization%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22William%20F.%22%2C%22lastName%22%3A%22Jenkins%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Yongsung%22%2C%22lastName%22%3A%22Park%22%7D%5D%2C%22abstractNote%22%3A%22Source%20localization%20with%20a%20geoacoustic%20model%20requires%20optimizing%20the%20model%20over%20a%20parameter%20space%20of%20range%20and%20depth%20with%20the%20objective%20of%20matching%20a%20predicted%20sound%20field%20to%20a%20field%20measured%20on%20an%20array.%20We%20propose%20a%20sample-efficient%20sequential%20Bayesian%20optimization%20strategy%20that%20models%20the%20objective%20function%20as%20a%20Gaussian%20process%20%28GP%29%20surrogate%20model%20conditioned%20on%20observed%20data.%20Using%20the%20mean%20and%20covariance%20functions%20of%20the%20GP%2C%20a%20heuristic%20acquisition%20function%20proposes%20a%20candidate%20in%20parameter%20space%20to%20sample%2C%20balancing%20exploitation%20%28sampling%20around%20the%20best%20observed%20objective%20function%20value%29%20and%20exploration%20%28sampling%20in%20regions%20of%20high%20variance%20in%20the%20GP%29.%20The%20candidate%20sample%20is%20evaluated%2C%20and%20the%20GP%20conditioned%20on%20the%20updated%20data.%20Optimization%20proceeds%20sequentially%20until%20a%20fixed%20budget%20of%20evaluations%20is%20expended.%20We%20demonstrate%20source%20localization%20for%20a%20shallow-water%20waveguide%20using%20Monte%20Carlo%20simulations%20and%20experimental%20data%20from%20an%20acoustic%20source%20tow.%20Compared%20to%20grid%20search%20and%20quasi-random%20sampling%20strategies%2C%20simulations%20and%20experimental%20results%20indicate%20the%20Bayesian%20optimization%20strategy%20converges%20on%20optimal%20solutions%20rapidly.%22%2C%22date%22%3A%222023-09-01%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0020839%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fpubs.aip.org%5C%2Fjasa%5C%2Farticle%5C%2F154%5C%2F3%5C%2F1459%5C%2F2909938%5C%2FBayesian-optimization-with-Gaussian-process%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%22A6U9IMQY%22%5D%2C%22dateModified%22%3A%222023-10-20T17%3A44%3A27Z%22%7D%7D%2C%7B%22key%22%3A%22AUX5MY4Z%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Khurjekar%20and%20Gerstoft%22%2C%22parsedDate%22%3A%222023-08-01%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EKhurjekar%2C%20I.%20D.%2C%20%26amp%3B%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%20%282023%29.%20Uncertainty%20quantification%20for%20direction-of-arrival%20estimation%20with%20conformal%20prediction.%20%3Ci%3EThe%20Journal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E154%3C%5C%2Fi%3E%282%29%2C%20979%26%23x2013%3B990.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0020655%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0020655%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Uncertainty%20quantification%20for%20direction-of-arrival%20estimation%20with%20conformal%20prediction%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Ishan%20D.%22%2C%22lastName%22%3A%22Khurjekar%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%5D%2C%22abstractNote%22%3A%22Uncertainty%20quantification%20%28UQ%29%20of%20deep%20learning%20%28DL%29-based%20acoustic%20estimation%20methods%20is%20useful%20for%20establishing%20confidence%20in%20the%20predictions.%20This%20is%20crucial%20to%20enable%20the%20real-world%20applicability%20of%20DL-based%20systems%20for%20acoustic%20tasks.%20Specifically%2C%20it%20is%20proposed%20to%20use%20conformal%20prediction%20%28CP%29%20for%20UQ%20in%20direction-of-arrival%20%28DOA%29%20estimation.%20CP%20is%20a%20statistically%20rigorous%20method%20to%20provide%20confidence%20intervals%20for%20an%20estimated%20quantity%20without%20making%20distributional%20assumptions.%20With%20CP%2C%20confidence%20intervals%20are%20computed%20via%20quantiles%20of%20user-defined%20scores.%20This%20easy-to-use%20method%20can%20be%20applied%20to%20any%20trained%20classification%5C%2Fregression%20model%20if%20an%20appropriate%20score%20function%20is%20chosen.%20The%20proposed%20approach%20shows%20the%20potential%20to%20enhance%20the%20real-time%20applicability%20of%20DL%20methods%20for%20DOA%20estimation.%20The%20advantages%20of%20CP%20are%20illustrated%20for%20different%20DL%20methods%20for%20DOA%20estimation%20in%20the%20presence%20of%20commonly%20occurring%20environmental%20uncertainty.%20Codes%20are%20available%20online%20%28https%3A%5C%2F%5C%2Fgithub.com%5C%2FNoiseLabUCSD%5C%2FConformalPrediction%29.%22%2C%22date%22%3A%222023-08-01%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0020655%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fpubs.aip.org%5C%2Fjasa%5C%2Farticle%5C%2F154%5C%2F2%5C%2F979%5C%2F2906684%5C%2FUncertainty-quantification-for-direction-of%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222023-09-14T18%3A53%3A39Z%22%7D%7D%2C%7B%22key%22%3A%22Z62VU9QC%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Lee%20et%20al.%22%2C%22parsedDate%22%3A%222023-07-01%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3ELee%2C%20J.-H.%2C%20Park%2C%20Y.%2C%20%26amp%3B%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%20%282023%29.%20Compressive%20frequency-difference%20direction-of-arrival%20estimation.%20%3Ci%3EThe%20Journal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E154%3C%5C%2Fi%3E%281%29%2C%20141%26%23x2013%3B151.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0020053%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0020053%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Compressive%20frequency-difference%20direction-of-arrival%20estimation%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Jeung-Hoon%22%2C%22lastName%22%3A%22Lee%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Yongsung%22%2C%22lastName%22%3A%22Park%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%5D%2C%22abstractNote%22%3A%22Direction-of-arrival%20estimation%20is%20difficult%20for%20signals%20spatially%20undersampled%20by%20more%20than%20half%20the%20wavelength.%20Frequency-difference%20beamforming%20%5BAbadi%2C%20Song%2C%20and%20Dowling%20%282012%29.%20J.%20Acoust.%20Soc.%20Am.%20132%2C%203018%5Cu20133029%5D%20offers%20an%20alternative%20approach%20to%20avoid%20such%20spatial%20aliasing%20by%20using%20multifrequency%20signals%20and%20processing%20them%20at%20a%20lower%20frequency%2C%20the%20difference-frequency.%20As%20with%20the%20conventional%20beamforming%20method%2C%20lowering%20the%20processing%20frequency%20sacrifices%20spatial%20resolution%20due%20to%20a%20beam%20broadening.%20Thus%2C%20unconventional%20beamforming%20is%20detrimental%20to%20the%20ability%20to%20distinguish%20between%20closely%20spaced%20targets.%20To%20overcome%20spatial%20resolution%20deterioration%2C%20we%20propose%20a%20simple%20yet%20effective%20method%20by%20formulating%20the%20frequency-difference%20beamforming%20as%20a%20sparse%20signal%20reconstruction%20problem.%20Similar%20to%20compressive%20beamforming%2C%20the%20improvement%20%28compressive%20frequency-difference%20beamforming%29%20promotes%20sparse%20nonzero%20elements%20to%20obtain%20a%20sharp%20estimate%20of%20the%20spatial%20direction-of-arrival%20spectrum.%20Analysis%20of%20the%20resolution%20limit%20demonstrates%20that%20the%20proposed%20method%20outperforms%20the%20conventional%20frequency-difference%20beamforming%20in%20terms%20of%20separation%20if%20the%20signal-to-noise%20ratio%20exceeds%204%5Cu2009dB.%20Ocean%20data%20from%20the%20FAF06%20experiment%20support%20the%20validity.%22%2C%22date%22%3A%222023-07-01%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0020053%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fpubs.aip.org%5C%2Fjasa%5C%2Farticle%5C%2F154%5C%2F1%5C%2F141%5C%2F2901954%5C%2FCompressive-frequency-difference-direction-of%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%22A6U9IMQY%22%5D%2C%22dateModified%22%3A%222023-08-13T00%3A31%3A58Z%22%7D%7D%2C%7B%22key%22%3A%22TJ9DVEM4%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Liu%20et%20al.%22%2C%22parsedDate%22%3A%222023-07-01%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3ELiu%2C%20Y.%2C%20Zhao%2C%20Y.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20Zhou%2C%20F.%2C%20Qiao%2C%20G.%2C%20%26amp%3B%20Yin%2C%20J.%20%282023%29.%20Deep%20transfer%20learning-based%20variable%20Doppler%20underwater%20acoustic%20communications.%20%3Ci%3EThe%20Journal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E154%3C%5C%2Fi%3E%281%29%2C%20232%26%23x2013%3B244.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0020147%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0020147%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Deep%20transfer%20learning-based%20variable%20Doppler%20underwater%20acoustic%20communications%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Yufei%22%2C%22lastName%22%3A%22Liu%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Yunjiang%22%2C%22lastName%22%3A%22Zhao%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Feng%22%2C%22lastName%22%3A%22Zhou%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Gang%22%2C%22lastName%22%3A%22Qiao%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Jingwei%22%2C%22lastName%22%3A%22Yin%22%7D%5D%2C%22abstractNote%22%3A%22This%20paper%20proposes%20a%20deep%20transfer%20learning%20%28DTL%29-based%20variable%20Doppler%20frequency-hopping%20binary%20frequency-shift%20keying%20underwater%20acoustic%20communication%20system.%20The%20system%20uses%20a%20convolutional%20neural%20network%20%28CNN%29%20as%20the%20demodulation%20module%20of%20the%20receiver.%20This%20approach%20directly%20demodulates%20the%20received%20signal%20without%20estimating%20the%20Doppler.%20The%20DTL%20first%20uses%20the%20simulated%20communication%20signal%20data%20to%20complete%20the%20CNN%20training.%20It%20then%20copies%20a%20part%20of%20the%20convolution%20layers%20from%20the%20pre-trained%20CNN%20to%20the%20target%20CNN.%20After%20randomly%20initializing%20the%20remaining%20layers%20for%20the%20target%20CNN%2C%20it%20is%20trained%20by%20the%20data%20samples%20from%20the%20specific%20communication%20scenarios.%20During%20the%20training%20process%2C%20the%20CNN%20learns%20the%20corresponding%20frequency%20from%20each%20symbol%20in%20the%20selected%20frequency-hopping%20group%20through%20the%20Mel-spectrograms.%20Simulation%20and%20experimental%20data%20processing%20results%20show%20that%20the%20performance%20of%20the%20proposed%20system%20is%20better%20than%20conventional%20systems%2C%20especially%20when%20the%20transmitter%20and%20receiver%20of%20the%20communication%20system%20are%20in%20variable%20speed%20motion%20in%20shallow%20water%20acoustic%20channels.%22%2C%22date%22%3A%222023-07-01%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0020147%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fpubs.aip.org%5C%2Fjasa%5C%2Farticle%5C%2F154%5C%2F1%5C%2F232%5C%2F2902195%5C%2FDeep-transfer-learning-based-variable-Doppler%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222023-08-13T00%3A33%3A58Z%22%7D%7D%2C%7B%22key%22%3A%22JXD77TK8%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Liu%20et%20al.%22%2C%22parsedDate%22%3A%222023-06-01%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3ELiu%2C%20R.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20Bianco%2C%20M.%20J.%2C%20%26amp%3B%20Rao%2C%20B.%20D.%20%282023%29.%20Recovery%20of%20spatially%20varying%20acoustical%20properties%20via%20automated%20partial%20differential%20equation%20identification.%20%3Ci%3EThe%20Journal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E153%3C%5C%2Fi%3E%286%29%2C%203169.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0019592%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0019592%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Recovery%20of%20spatially%20varying%20acoustical%20properties%20via%20automated%20partial%20differential%20equation%20identification%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Ruixian%22%2C%22lastName%22%3A%22Liu%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Michael%20J%22%2C%22lastName%22%3A%22Bianco%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Bhaskar%20D.%22%2C%22lastName%22%3A%22Rao%22%7D%5D%2C%22abstractNote%22%3A%22Observable%20dynamics%2C%20such%20as%20waves%20propagating%20on%20a%20surface%2C%20are%20generally%20governed%20by%20partial%20differential%20equations%20%28PDEs%29%2C%20which%20are%20determined%20by%20the%20physical%20properties%20of%20the%20propagation%20media.%20The%20spatial%20variations%20of%20these%20properties%20lead%20to%20spatially%20dependent%20PDEs.%20It%20is%20useful%20in%20many%20fields%20to%20recover%20the%20variations%20from%20the%20observations%20of%20dynamical%20behaviors%20on%20the%20material.%20A%20method%20is%20proposed%20to%20form%20a%20map%20of%20the%20physical%20properties%27%20spatial%20variations%20for%20a%20material%20via%20data-driven%20spatially%20dependent%20PDE%20identification%20and%20applied%20to%20recover%20acoustical%20properties%20%28viscosity%2C%20attenuation%2C%20and%20phase%20speeds%29%20for%20propagating%20waves.%20The%20proposed%20data-driven%20PDE%20identification%20scheme%20is%20based%20on%20%5Cu21131-norm%20minimization.%20It%20does%20not%20require%20any%20PDE%20term%20that%20is%20assumed%20active%20from%20the%20prior%20knowledge%20and%20is%20the%20first%20approach%20that%20is%20capable%20of%20identifying%20spatially%20dependent%20PDEs%20from%20measurements%20of%20phenomena.%20In%20addition%2C%20the%20method%20is%20efficient%20as%20a%20result%20of%20its%20non-iterative%20nature%20and%20can%20be%20robust%20against%20noise%20if%20used%20with%20an%20integration%20transformation%20technique.%20It%20is%20demonstrated%20in%20multiple%20experimental%20settings%2C%20including%20real%20laser%20measurements%20of%20a%20vibrating%20aluminum%20plate.%20Codes%20and%20data%20are%20available%20online%20at%20https%3A%5C%2F%5C%2Ftinyurl.com%5C%2F4wza8vxs.%22%2C%22date%22%3A%222023-06-01%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0019592%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fpubs.aip.org%5C%2Fjasa%5C%2Farticle%5C%2F153%5C%2F6%5C%2F3169%5C%2F2894384%5C%2FRecovery-of-spatially-varying-acoustical%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222023-07-17T22%3A37%3A55Z%22%7D%7D%2C%7B%22key%22%3A%22PNMZ56T2%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Michalopoulou%20and%20Gerstoft%22%2C%22parsedDate%22%3A%222023-03%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EMichalopoulou%2C%20Z.%20H.%2C%20%26amp%3B%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%20%282023%29.%20Waveform%20modeling%20with%20Gaussian%20Processes%20for%20inversion%20in%20ocean%20acoustics.%20%3Ci%3EJournal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E153%3C%5C%2Fi%3E%283%29.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0018577%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0018577%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Waveform%20modeling%20with%20Gaussian%20Processes%20for%20inversion%20in%20ocean%20acoustics%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Z.%20H.%22%2C%22lastName%22%3A%22Michalopoulou%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%5D%2C%22abstractNote%22%3A%22%22%2C%22date%22%3A%22Mar%202023%22%2C%22language%22%3A%22English%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0018577%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%2289L3J9FA%22%5D%2C%22dateModified%22%3A%222024-05-03T16%3A55%3A54Z%22%7D%7D%2C%7B%22key%22%3A%224UKH6XGK%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Michalopoulou%20and%20Gerstoft%22%2C%22parsedDate%22%3A%222023-03%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EMichalopoulou%2C%20Z.%20H.%2C%20%26amp%3B%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%20%282023%29.%20Inversion%20in%20an%20uncertain%20ocean%20using%20Gaussian%20processes.%20%3Ci%3EJournal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E153%3C%5C%2Fi%3E%283%29%2C%201600%26%23x2013%3B1611.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0017437%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0017437%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Inversion%20in%20an%20uncertain%20ocean%20using%20Gaussian%20processes%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Z.%20H.%22%2C%22lastName%22%3A%22Michalopoulou%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%5D%2C%22abstractNote%22%3A%22Gaussian%20processes%20%28GPs%29%20can%20capture%20correlation%20of%20the%20acoustic%20field%20at%20different%20depths%20in%20the%20ocean.%20This%20feature%20is%20exploited%20in%20this%20work%20for%20pre-processing%20acoustic%20data%20before%20these%20are%20employed%20for%20source%20localization%20and%20environmental%20inversion%20using%20matched%20field%20inversion%20%28MFI%29%20in%20an%20underwater%20waveguide.%20Via%20the%20application%20of%20GPs%2C%20the%20data%20are%20denoised%20and%20interpolated%2C%20generating%20densely%20populated%20acoustic%20fields%20at%20virtual%20arrays%2C%20which%20are%20then%20used%20as%20data%20in%20MFI.%20Replicas%20are%20also%20computed%20at%20the%20virtual%20receivers%20at%20which%20field%20predictions%20are%20made.%20The%20correlations%20among%20field%20measurements%20at%20distinct%20spatial%20points%20are%20manifested%20through%20the%20selection%20of%20kernel%20functions.%20These%20rely%20on%20hyperparameters%2C%20that%20are%20estimated%20through%20a%20maximum%20likelihood%20process%20for%20optimal%20denoising%20and%20interpolation.%20The%20approach%2C%20employing%20Gaussian%20and%20Matern%20kernels%2C%20is%20tested%20on%20synthetic%20and%20real%20data%20with%20both%20an%20exhaustive%20search%20and%20genetic%20algorithms%20and%20is%20found%20to%20be%20superior%20to%20conventional%20beamformer%20MFI.%20It%20is%20also%20shown%20that%20the%20Matern%20kernel%2C%20providing%20more%20degrees%20of%20freedom%20because%20of%20an%20increased%20number%20of%20hyperparameters%2C%20is%20preferable%20over%20the%20frequently%20used%20Gaussian%20kernel.%22%2C%22date%22%3A%22Mar%202023%22%2C%22language%22%3A%22English%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0017437%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%2289L3J9FA%22%5D%2C%22dateModified%22%3A%222024-05-03T16%3A55%3A54Z%22%7D%7D%2C%7B%22key%22%3A%223QKJK5Q5%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Jenkins%20et%20al.%22%2C%22parsedDate%22%3A%222023-03%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EJenkins%2C%20W.%20F.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20Chien%2C%20C.%20C.%2C%20%26amp%3B%20Ozanich%2C%20E.%20%282023%29.%20Reducing%20dimensionality%20of%20spectrograms%20using%20convolutional%20autoencoders.%20%3Ci%3EJournal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E153%3C%5C%2Fi%3E%283%29.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0018582%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0018582%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Reducing%20dimensionality%20of%20spectrograms%20using%20convolutional%20autoencoders%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22W.%20F.%22%2C%22lastName%22%3A%22Jenkins%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22C.%20C.%22%2C%22lastName%22%3A%22Chien%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22E.%22%2C%22lastName%22%3A%22Ozanich%22%7D%5D%2C%22abstractNote%22%3A%22%22%2C%22date%22%3A%22Mar%202023%22%2C%22language%22%3A%22English%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0018582%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%2289L3J9FA%22%5D%2C%22dateModified%22%3A%222024-05-03T16%3A55%3A54Z%22%7D%7D%2C%7B%22key%22%3A%222SEGZ66S%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Chien%20et%20al.%22%2C%22parsedDate%22%3A%222023%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EChien%2C%20C.-C.%2C%20Jenkins%2C%20W.%20F.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20Zumberge%2C%20M.%2C%20%26amp%3B%20Mellors%2C%20R.%20%282023%29.%20Automatic%20classification%20with%20an%20autoencoder%20of%20seismic%20signals%20on%20a%20distributed%20acoustic%20sensing%20cable.%20%3Ci%3EComputers%20and%20Geotechnics%3C%5C%2Fi%3E%2C%20%3Ci%3E155%3C%5C%2Fi%3E%2C%20105223.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1016%5C%2Fj.compgeo.2022.105223%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1016%5C%2Fj.compgeo.2022.105223%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Automatic%20classification%20with%20an%20autoencoder%20of%20seismic%20signals%20on%20a%20distributed%20acoustic%20sensing%20cable%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Chih-Chieh%22%2C%22lastName%22%3A%22Chien%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22William%20F.%22%2C%22lastName%22%3A%22Jenkins%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Mark%22%2C%22lastName%22%3A%22Zumberge%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Robert%22%2C%22lastName%22%3A%22Mellors%22%7D%5D%2C%22abstractNote%22%3A%22%22%2C%22date%22%3A%2203%5C%2F2023%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1016%5C%2Fj.compgeo.2022.105223%22%2C%22ISSN%22%3A%220266352X%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Flinkinghub.elsevier.com%5C%2Fretrieve%5C%2Fpii%5C%2FS0266352X22005602%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%22P7USRYNU%22%5D%2C%22dateModified%22%3A%222023-03-10T21%3A41%3A38Z%22%7D%7D%2C%7B%22key%22%3A%22F8YC95NN%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Fernandez-Grande%20et%20al.%22%2C%22parsedDate%22%3A%222023%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EFernandez-Grande%2C%20E.%2C%20Karakonstantis%2C%20X.%2C%20Caviedes-Nozal%2C%20D.%2C%20%26amp%3B%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%20%282023%29.%20Generative%20models%20for%20sound%20field%20reconstruction.%20%3Ci%3EThe%20Journal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E153%3C%5C%2Fi%3E%282%29%2C%201179%26%23x2013%3B1190.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0016896%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0016896%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Generative%20models%20for%20sound%20field%20reconstruction%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Efren%22%2C%22lastName%22%3A%22Fernandez-Grande%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Xenofon%22%2C%22lastName%22%3A%22Karakonstantis%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Diego%22%2C%22lastName%22%3A%22Caviedes-Nozal%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%5D%2C%22abstractNote%22%3A%22This%20work%20examines%20the%20use%20of%20generative%20adversarial%20networks%20for%20reconstructing%20sound%20fields%20from%20experimental%20data.%20It%20is%20investigated%20whether%20generative%20models%2C%20which%20learn%20the%20underlying%20statistics%20of%20a%20given%20signal%20or%20process%2C%20can%20improve%20the%20spatio-temporal%20reconstruction%20of%20a%20sound%20field%20by%20extending%20its%20bandwidth.%20The%20problem%20is%20significant%20as%20acoustic%20array%20processing%20is%20naturally%20band%20limited%20by%20the%20spatial%20sampling%20of%20the%20sound%20field%20%28due%20to%20the%20difficulty%20to%20satisfy%20the%20Nyquist%20criterion%20in%20space%20domain%20at%20high%20frequencies%29.%20In%20this%20study%2C%20the%20reconstruction%20of%20spatial%20room%20impulse%20responses%20in%20a%20conventional%20room%20is%20tested%20based%20on%20three%20different%20generative%20adversarial%20models.%20The%20results%20indicate%20that%20the%20models%20can%20improve%20the%20reconstruction%2C%20mostly%20by%20recovering%20some%20of%20the%20sound%20field%20energy%20that%20would%20otherwise%20be%20lost%20at%20high%20frequencies.%20There%20is%20an%20encouraging%20outlook%20in%20the%20use%20of%20statistical%20learning%20models%20to%20overcome%20the%20bandwidth%20limitations%20of%20acoustic%20sensor%20arrays.%20The%20approach%20can%20be%20of%20interest%20in%20other%20areas%2C%20such%20as%20computational%20acoustics%2C%20to%20alleviate%20the%20classical%20computational%20burden%20at%20high%20frequencies.%22%2C%22date%22%3A%2202%5C%2F2023%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0016896%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fasa.scitation.org%5C%2Fdoi%5C%2F10.1121%5C%2F10.0016896%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222023-04-11T21%3A24%3A46Z%22%7D%7D%2C%7B%22key%22%3A%22R4WCKFPS%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Park%20et%20al.%22%2C%22parsedDate%22%3A%222023%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EPark%2C%20Y.%2C%20Meyer%2C%20F.%2C%20%26amp%3B%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%20%282023%29.%20Graph-based%20sequential%20beamforming.%20%3Ci%3EThe%20Journal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E153%3C%5C%2Fi%3E%281%29%2C%20723%26%23x2013%3B737.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0016876%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0016876%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Graph-based%20sequential%20beamforming%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Yongsung%22%2C%22lastName%22%3A%22Park%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Florian%22%2C%22lastName%22%3A%22Meyer%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%5D%2C%22abstractNote%22%3A%22This%20paper%20presents%20a%20Bayesian%20estimation%20method%20for%20sequential%20direction%20finding.%20The%20proposed%20method%20estimates%20the%20number%20of%20directions%20of%20arrivals%20%28DOAs%29%20and%20their%20DOAs%20performing%20operations%20on%20the%20factor%20graph.%20The%20graph%20represents%20a%20statistical%20model%20for%20sequential%20beamforming.%20At%20each%20time%20step%2C%20belief%20propagation%20predicts%20the%20number%20of%20DOAs%20and%20their%20DOAs%20using%20posterior%20probability%20density%20functions%20%28pdfs%29%20from%20the%20previous%20time%20and%20a%20different%20Bernoulli-von%20Mises%20state%20transition%20model.%20Variational%20Bayesian%20inference%20then%20updates%20the%20number%20of%20DOAs%20and%20their%20DOAs.%20The%20method%20promotes%20sparse%20solutions%20through%20a%20Bernoulli-Gaussian%20amplitude%20model%2C%20is%20gridless%2C%20and%20provides%20marginal%20posterior%20pdfs%20from%20which%20DOA%20estimates%20and%20their%20uncertainties%20can%20be%20extracted.%20Compared%20to%20nonsequential%20approaches%2C%20the%20method%20can%20reduce%20DOA%20estimation%20errors%20in%20scenarios%20involving%20multiple%20time%20steps%20and%20time-varying%20DOAs.%20Simulation%20results%20demonstrate%20performance%20improvements%20compared%20to%20state-of-the-art%20methods.%20The%20proposed%20method%20is%20evaluated%20using%20ocean%20acoustic%20experimental%20data.%22%2C%22date%22%3A%2201%5C%2F2023%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0016876%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fasa.scitation.org%5C%2Fdoi%5C%2F10.1121%5C%2F10.0016876%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%22764DPZJV%22%2C%22A6U9IMQY%22%5D%2C%22dateModified%22%3A%222023-03-20T15%3A26%3A24Z%22%7D%7D%2C%7B%22key%22%3A%2293Y7EUBY%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22De%20Salvio%20et%20al.%22%2C%22parsedDate%22%3A%222023%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EDe%20Salvio%2C%20D.%2C%20Bianco%2C%20M.%20J.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20D%26%23x2019%3BOrazio%2C%20D.%2C%20%26amp%3B%20Garai%2C%20M.%20%282023%29.%20Blind%20source%20separation%20by%20long-term%20monitoring%3A%20A%20variational%20autoencoder%20to%20validate%20the%20clustering%20analysis.%20%3Ci%3EThe%20Journal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E153%3C%5C%2Fi%3E%281%29%2C%20738%26%23x2013%3B750.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0016887%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0016887%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Blind%20source%20separation%20by%20long-term%20monitoring%3A%20A%20variational%20autoencoder%20to%20validate%20the%20clustering%20analysis%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Domenico%22%2C%22lastName%22%3A%22De%20Salvio%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Michael%20J.%22%2C%22lastName%22%3A%22Bianco%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Dario%22%2C%22lastName%22%3A%22D%27Orazio%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Massimo%22%2C%22lastName%22%3A%22Garai%22%7D%5D%2C%22abstractNote%22%3A%22Noise%20exposure%20influences%20the%20comfort%20and%20well-being%20of%20people%20in%20several%20contexts%2C%20such%20as%20work%20or%20learning%20environments.%20For%20instance%2C%20in%20offices%2C%20different%20kind%20of%20noises%20can%20increase%20or%20drop%20the%20employees%27%20productivity.%20Thus%2C%20the%20ability%20of%20separating%20sound%20sources%20in%20real%20contexts%20plays%20a%20key%20role%20in%20assessing%20sound%20environments.%20Long-term%20monitoring%20provide%20large%20amounts%20of%20data%20that%20can%20be%20analyzed%20through%20machine%20and%20deep%20learning%20algorithms.%20Based%20on%20previous%20works%2C%20an%20entire%20working%20day%20was%20recorded%20through%20a%20sound%20level%20meter.%20Both%20sound%20pressure%20levels%20and%20the%20digital%20audio%20recording%20were%20collected.%20Then%2C%20a%20dual%20clustering%20analysis%20was%20carried%20out%20to%20separate%20the%20two%20main%20sound%20sources%20experienced%20by%20workers%3A%20traffic%20and%20speech%20noises.%20The%20first%20method%20exploited%20the%20occurrences%20of%20sound%20pressure%20levels%20via%20Gaussian%20mixture%20model%20and%20K-means%20clustering.%20The%20second%20analysis%20performed%20a%20semi-supervised%20deep%20clustering%20analyzing%20the%20latent%20space%20of%20a%20variational%20autoencoder.%20Results%20show%20that%20both%20approaches%20were%20able%20to%20separate%20the%20sound%20sources.%20Spectral%20matching%20and%20the%20latent%20space%20of%20the%20variational%20autoencoder%20validated%20the%20assumptions%20underlying%20the%20proposed%20clustering%20methods.%22%2C%22date%22%3A%2201%5C%2F2023%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0016887%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fasa.scitation.org%5C%2Fdoi%5C%2F10.1121%5C%2F10.0016887%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222023-03-10T21%3A43%3A12Z%22%7D%7D%2C%7B%22key%22%3A%22P2SWIHNU%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Wu%20et%20al.%22%2C%22parsedDate%22%3A%222023%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EWu%2C%20Y.%20F.%2C%20Wakin%2C%20M.%20B.%2C%20%26amp%3B%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%20%282023%29.%20Gridless%20DOA%20Estimation%20With%20Multiple%20Frequencies.%20%3Ci%3EIeee%20Transactions%20on%20Signal%20Processing%3C%5C%2Fi%3E%2C%20%3Ci%3E71%3C%5C%2Fi%3E%2C%20417%26%23x2013%3B432.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2FTsp.2023.3244091%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2FTsp.2023.3244091%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Gridless%20DOA%20Estimation%20With%20Multiple%20Frequencies%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Y.%20F.%22%2C%22lastName%22%3A%22Wu%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22M.%20B.%22%2C%22lastName%22%3A%22Wakin%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%5D%2C%22abstractNote%22%3A%22Direction-of-arrival%20%28DOA%29%20estimation%20is%20widely%20applied%20in%20acoustic%20source%20localization.%20A%20multi-frequency%20model%20is%20suitable%20for%20characterizing%20the%20broadband%20structure%20in%20acoustic%20signals.%20In%20this%20paper%2C%20the%20continuous%20%28gridless%29%20DOA%20estimation%20problem%20with%20multiple%20frequencies%20is%20considered.%20This%20problem%20is%20formulated%20as%20an%20atomic%20norm%20minimization%20%28ANM%29%20problem.%20The%20ANM%20problem%20is%20equivalent%20to%20a%20semi-definite%20program%20%28SDP%29%20which%20can%20be%20solved%20by%20an%20off-the-shelf%20SDP%20solver.%20The%20dual%20certificate%20condition%20is%20provided%20to%20certify%20the%20optimality%20of%20the%20SDP%20solution%20so%20that%20the%20sources%20can%20be%20localized%20by%20finding%20the%20roots%20of%20a%20polynomial.%20We%20also%20construct%20the%20dual%20polynomial%20to%20satisfy%20the%20dual%20certificate%20condition%20and%20show%20that%20such%20a%20construction%20exists%20when%20the%20source%20amplitude%20has%20a%20uniform%20magnitude.%20In%20multi-frequency%20ANM%2C%20spatial%20aliasing%20of%20DOAs%20at%20higher%20frequencies%20can%20cause%20challenges.%20We%20discuss%20this%20issue%20extensively%20and%20propose%20a%20robust%20solution%20to%20combat%20aliasing.%20Numerical%20results%20support%20our%20theoretical%20findings%20and%20demonstrate%20the%20effectiveness%20of%20the%20proposed%20method.%22%2C%22date%22%3A%222023%22%2C%22language%22%3A%22English%22%2C%22DOI%22%3A%2210.1109%5C%2FTsp.2023.3244091%22%2C%22ISSN%22%3A%221053-587x%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%2289L3J9FA%22%5D%2C%22dateModified%22%3A%222024-05-03T16%3A55%3A54Z%22%7D%7D%2C%7B%22key%22%3A%22HYUYPP7N%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Chaput%20et%20al.%22%2C%22parsedDate%22%3A%222022-12-19%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EChaput%2C%20J.%2C%20Aster%2C%20R.%2C%20Karplus%2C%20M.%2C%20Nakata%2C%20N.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20Bromirski%2C%20P.%20D.%2C%20Nyblade%2C%20A.%2C%20Stephen%2C%20R.%20A.%2C%20%26amp%3B%20Wiens%2C%20D.%20A.%20%282022%29.%20Near-surface%20seismic%20anisotropy%20in%20Antarctic%20glacial%20snow%20and%20ice%20revealed%20by%20high-frequency%20ambient%20noise.%20%3Ci%3EJournal%20of%20Glaciology%3C%5C%2Fi%3E%2C%201%26%23x2013%3B17.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1017%5C%2Fjog.2022.98%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1017%5C%2Fjog.2022.98%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Near-surface%20seismic%20anisotropy%20in%20Antarctic%20glacial%20snow%20and%20ice%20revealed%20by%20high-frequency%20ambient%20noise%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Julien%22%2C%22lastName%22%3A%22Chaput%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Rick%22%2C%22lastName%22%3A%22Aster%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Marianne%22%2C%22lastName%22%3A%22Karplus%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Nori%22%2C%22lastName%22%3A%22Nakata%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%20D.%22%2C%22lastName%22%3A%22Bromirski%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22A.%22%2C%22lastName%22%3A%22Nyblade%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22R.%20A.%22%2C%22lastName%22%3A%22Stephen%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22D.%20A.%22%2C%22lastName%22%3A%22Wiens%22%7D%5D%2C%22abstractNote%22%3A%22Abstract%5Cn%20%20%20%20%20%20%20%20%20%20%20%20Ambient%20seismic%20recordings%20taken%20at%20broad%20locations%20across%20Ross%20Ice%20Shelf%20and%20a%20dense%20array%20near%20West%20Antarctic%20Ice%20Sheet%20%28WAIS%29%20Divide%2C%20Antarctica%2C%20show%20pervasive%20temporally%20variable%20resonance%20peaks%20associated%20with%20trapped%20seismic%20waves%20in%20near-surface%20firn%20layers.%20These%20resonance%20peaks%20feature%20splitting%20on%20the%20horizontal%20components%2C%20here%20interpreted%20as%20frequency-dependent%20anisotropy%20in%20the%20firn%20and%20underlying%20ice%20due%20to%20several%20overlapping%20mechanisms%20driven%20by%20ice%20flow.%20Frequency%20peak%20splitting%20magnitudes%20and%20fast%5C%2Fslow%20axes%20were%20systematically%20estimated%20at%20single%20stations%20using%20a%20novel%20algorithm%20and%20compared%20with%20good%20agreement%20with%20active%20source%20anisotropy%20measurements%20at%20WAIS%20Divide%20determined%20via%20active%20sources%20recorded%20on%20a%201%20km%20circular%20array.%20The%20approach%20was%20further%20applied%20to%20the%20broad%20Ross%20Ice%20Shelf%20%28RIS%29%20array%2C%20where%20anisotropy%20axes%20were%20directly%20compared%20with%20visible%20surface%20features%20and%20ice%20shelf%20flow%20lines.%20The%20near-surface%20firn%2C%20depicted%20by%20anisotropy%20above%2030%20Hz%2C%20was%20shown%20to%20exhibit%20a%20novel%20plastic%20stretching%20mechanism%20of%20anisotropy%2C%20whereby%20the%20fast%20direction%20in%20snow%20aligns%20with%20accelerating%20ice%20shelf%20flow.%22%2C%22date%22%3A%222022-12-19%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1017%5C%2Fjog.2022.98%22%2C%22ISSN%22%3A%220022-1430%2C%201727-5652%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fwww.cambridge.org%5C%2Fcore%5C%2Fproduct%5C%2Fidentifier%5C%2FS0022143022000983%5C%2Ftype%5C%2Fjournal_article%22%2C%22collections%22%3A%5B%22D2UAFEX5%22%2C%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222023-01-19T21%3A50%3A07Z%22%7D%7D%2C%7B%22key%22%3A%22PXPDZYX2%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Collins%20et%20al.%22%2C%22parsedDate%22%3A%222022-09%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3ECollins%2C%20M.%20D.%2C%20Turgut%2C%20A.%2C%20Buckingham%2C%20M.%20J.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20%26amp%3B%20Siderius%2C%20M.%20%282022%29.%20Selected%20Topics%20of%20the%20Past%20Thirty%20Years%20in%20Ocean%20Acoustics.%20%3Ci%3EJournal%20of%20Theoretical%20and%20Computational%20Acoustics%3C%5C%2Fi%3E%2C%20%3Ci%3E30%3C%5C%2Fi%3E%2803%29%2C%2022.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1142%5C%2Fs2591728522400011%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1142%5C%2Fs2591728522400011%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Selected%20Topics%20of%20the%20Past%20Thirty%20Years%20in%20Ocean%20Acoustics%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22M.%20D.%22%2C%22lastName%22%3A%22Collins%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22A.%22%2C%22lastName%22%3A%22Turgut%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22M.%20J.%22%2C%22lastName%22%3A%22Buckingham%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22M.%22%2C%22lastName%22%3A%22Siderius%22%7D%5D%2C%22abstractNote%22%3A%22This%20paper%20reviews%20some%20of%20the%20highlights%20of%20selected%20topics%20in%20ocean%20acoustics%20during%20the%20thirty%20years%20that%20have%20passed%20since%20the%20founding%20of%20the%20Journal%20of%20Theoretical%20and%20Computational%20Acoustics.%20Advances%20in%20computational%20methods%20and%20computers%20helped%20to%20make%20computational%20ocean%20acoustics%20a%20vibrant%20area%20of%20research%20during%20that%20period.%20The%20parabolic%20equation%20method%20provides%20an%20unrivaled%20combination%20of%20accuracy%20and%20efficiency%20for%20propagation%20problems%20in%20which%20the%20bathymetry%2C%20sound%20speed%2C%20and%20other%20environmental%20parameters%20vary%20in%20the%20horizontal%20directions.%20The%20extension%20of%20this%20approach%20to%20cases%20involving%20layers%20that%20support%20shear%20waves%20has%20been%20an%20active%20area%20of%20research%20throughout%20the%20thirty%20year%20period.%20Interest%20in%20basin-scale%20and%20global-scale%20propagation%20was%20stimulated%20by%20the%20Heard%20Island%20Feasibility%20Test%20for%20monitoring%20climate%20change%20in%20terms%20of%20changes%20in%20travel%20time%20that%20occur%20as%20the%20temperature%20of%20the%20ocean%20rises.%20Diminishing%20ice%20cover%20in%20the%20Arctic%2C%20which%20is%20one%20of%20the%20consequences%20of%20climate%20change%2C%20has%20stimulated%20renewed%20interest%20in%20Arctic%20acoustics%20during%20the%20past%20decade.%20Reverberation%20is%20a%20challenging%20problem%20that%20was%20the%20topic%20of%20a%20major%20research%20program%20during%20the%20beginning%20of%20the%20thirty%20year%20period.%20An%20innovative%20approach%20for%20making%20it%20feasible%20to%20solve%20such%20problems%20was%20applied%20to%20data%20for%20reverberation%20from%20the%20seafloor%20and%20from%20schools%20of%20fish%2C%20and%20some%20of%20the%20findings%20were%20featured%20in%20Science%20and%20Nature.%20Source%20localization%20is%20one%20of%20the%20core%20problems%20in%20ocean%20acoustics.%20When%20applied%20on%20a%202-D%20array%20of%20receivers%2C%20an%20approach%20based%20on%20the%20eigenvectors%20of%20the%20covariance%20matrix%20is%20capable%20of%20separating%20the%20signals%20from%20different%20sources%20from%20each%20other%2C%20determining%20when%20this%20partitioning%20step%20is%20successful%2C%20and%20tracking%20sources%20that%20cross%20each%20other%20in%20bearing%3B%20one%20of%20the%20advantages%20of%20this%20approach%20is%20that%20it%20does%20not%20require%20environmental%20information%20or%20solutions%20of%20the%20wave%20equation.%20Geoacoustic%20inversion%20for%20estimating%20the%20layer%20structure%2C%20wave%20speeds%2C%20density%2C%20and%20other%20parameters%20of%20ocean%20bottoms%20has%20also%20been%20a%20topic%20of%20interest%20throughout%20the%20thirty%20year%20period.%22%2C%22date%22%3A%22Sep%202022%22%2C%22language%22%3A%22English%22%2C%22DOI%22%3A%2210.1142%5C%2Fs2591728522400011%22%2C%22ISSN%22%3A%222591-7285%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22WICVHEJF%22%2C%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222022-11-21T21%3A34%3A29Z%22%7D%7D%2C%7B%22key%22%3A%227EXRW5NU%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Michalopoulou%20et%20al.%22%2C%22parsedDate%22%3A%222022-07%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EMichalopoulou%2C%20Z.%20H.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20Rios%2C%20D.%2C%20%26amp%3B%20Hodgkiss%2C%20W.%20S.%20%282022%29.%20Tracking%20and%20Inversion%20Using%20Midfrequency%20Signals%20in%20the%20Seabed%20Characterization%20Experiment.%20%3Ci%3EIeee%20Journal%20of%20Oceanic%20Engineering%3C%5C%2Fi%3E%2C%20%3Ci%3E47%3C%5C%2Fi%3E%283%29%2C%20657%26%23x2013%3B669.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2FJoe.2021.3122284%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2FJoe.2021.3122284%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Tracking%20and%20Inversion%20Using%20Midfrequency%20Signals%20in%20the%20Seabed%20Characterization%20Experiment%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Z.%20H.%22%2C%22lastName%22%3A%22Michalopoulou%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22D.%22%2C%22lastName%22%3A%22Rios%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22W.%20S.%22%2C%22lastName%22%3A%22Hodgkiss%22%7D%5D%2C%22abstractNote%22%3A%22Multipath%20arrivals%20are%20identified%20and%20extracted%20from%20midfrequency%20impulse%20responses%20at%20a%20vertical%20line%20array%20produced%20by%20a%20broadband%20source%20transmitting%20linearly%20frequency%20modulated%20pulses%20in%20the%20SBCEX%2017%20experiment.%20Arrival%20times%20are%20extracted%20from%20received%20data%20using%20particle%20filtering%20and%20smoothing.%20Specifically%2C%20sequential%20filtering%20is%20applied%20to%20the%20received%20time%20series%20at%20sixteen%20phones%20for%20the%20estimation%20of%20three%20paths%20generated%20by%20sound%20interacting%20with%20the%20propagation%20medium.%20This%20is%20followed%20up%20with%20a%20smoother%20that%20provides%20%5C%22crisp%5C%22%20arrival%20time%20probability%20densities.%20A%20peak-finding%20method%20is%20used%20to%20identify%20a%20fourth%20path.%20The%20estimated%20arrival%20times%20obtained%20by%20the%20process%20above%20are%20linked%20with%20a%20sound%20propagation%20model%20based%20on%20ray%20tracing%20and%20linearization%20for%20estimation%20of%20source%20location%20and%20water%20column%20depth%20and%20sound%20speed.%20An%20exhaustive%20search%20follows%20for%20inversion%20for%20sediment%20sound%20speed%20and%20thickness.%20Because%20the%20source%20is%20moving%2C%20receptions%20are%20available%20on%20a%20track%20and%20arrival%20time%20estimation%20and%20corresponding%20inversion%20are%20performed%20at%20every%20source%20location%20within%20the%20track.%20The%20unknown%20parameters%20are%20estimated%20sequentially%20with%20two%20approaches.%20Approach%20one%20is%20linearization%20using%20prior%20information%20from%20one%20location%20propagated%20to%20the%20next.%20Approach%20two%20relies%20on%20a%20second%20particle%20filter%2C%20capturing%20the%20dynamic%20evolution%20of%20parameters%20along%20the%20track.%20Both%20techniques%20lead%20to%20sediment%20property%20estimation.%20Results%20from%20the%20two%20methods%20agree%2C%20pointing%20to%20robustness%20in%20inversion%20when%20impulse%20responses%20and%20arrival%20times%20estimated%20from%20those%20are%20employed%20as%20input%20to%20the%20inverse%20problem.%22%2C%22date%22%3A%22Jul%202022%22%2C%22language%22%3A%22English%22%2C%22DOI%22%3A%2210.1109%5C%2FJoe.2021.3122284%22%2C%22ISSN%22%3A%220364-9059%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%222AQ2ECNM%22%2C%2289L3J9FA%22%5D%2C%22dateModified%22%3A%222024-05-03T16%3A55%3A54Z%22%7D%7D%2C%7B%22key%22%3A%22KGJCV3GW%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Park%20and%20Gerstoft%22%2C%22parsedDate%22%3A%222022-06%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EPark%2C%20Y.%2C%20%26amp%3B%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%20%282022%29.%20Gridless%20sparse%20covariance-based%20beamforming%20via%20alternating%20projections%20including%20co-prime%20arrays.%20%3Ci%3EJournal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E151%3C%5C%2Fi%3E%286%29%2C%203828%26%23x2013%3B3837.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0011617%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0011617%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Gridless%20sparse%20covariance-based%20beamforming%20via%20alternating%20projections%20including%20co-prime%20arrays%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Y.%22%2C%22lastName%22%3A%22Park%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%5D%2C%22abstractNote%22%3A%22This%20paper%20presents%20gridless%20sparse%20processing%20for%20direction-of-arrival%20%28DOA%29%20estimation.%20The%20method%20solves%20a%20gridless%20version%20of%20sparse%20covariance-based%20estimation%20using%20alternating%20projections.%20Gridless%20sparse%20DOA%20estimation%20is%20represented%20by%20the%20reconstruction%20of%20Toeplitz-structured%20low-rank%20matrices%2C%20which%20our%20method%20recovers%20by%20alternatively%20projecting%20a%20solution%20matrix.%20Compared%20to%20the%20existing%20gridless%20sparse%20methods%2C%20our%20method%20improves%20speed%20and%20accuracy%20and%20considers%20non-uniformly%20configured%20linear%20arrays.%20High-resolution%20and%20reliable%20DOA%20estimation%20are%20achieved%20even%20with%20single-snapshot%20data%2C%20coherent%20sources%2C%20and%20non-uniform%20arrays.%20Simulation%20results%20demonstrate%20performance%20improvements%20compared%20to%20the%20existing%20DOA%20estimators%2C%20including%20gridless%20sparse%20methods.%20The%20method%20is%20illustrated%20using%20experimental%20data%20from%20a%20real%20ocean%20experiment.%20%28C%29%202022%20Acoustical%20Society%20of%20America.%22%2C%22date%22%3A%222022%5C%2F06%22%2C%22language%22%3A%22English%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0011617%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%22A6U9IMQY%22%5D%2C%22dateModified%22%3A%222022-07-13T16%3A48%3A18Z%22%7D%7D%2C%7B%22key%22%3A%22LUBGDNAW%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Gerstoft%20et%20al.%22%2C%22parsedDate%22%3A%222022-03%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3E%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20Hu%2C%20Y.%20H.%2C%20Bianco%2C%20M.%20J.%2C%20Patil%2C%20C.%2C%20Alegre%2C%20A.%2C%20Freund%2C%20Y.%2C%20%26amp%3B%20Grondin%2C%20F.%20%282022%29.%20Audio%20scene%20monitoring%20using%20redundant%20ad%20hoc%20microphone%20array%20networks.%20%3Ci%3EIeee%20Internet%20of%20Things%20Journal%3C%5C%2Fi%3E%2C%20%3Ci%3E9%3C%5C%2Fi%3E%286%29%2C%204259%26%23x2013%3B4268.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2Fjiot.2021.3103523%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2Fjiot.2021.3103523%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Audio%20scene%20monitoring%20using%20redundant%20ad%20hoc%20microphone%20array%20networks%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Y.%20H.%22%2C%22lastName%22%3A%22Hu%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22M.%20J.%22%2C%22lastName%22%3A%22Bianco%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22C.%22%2C%22lastName%22%3A%22Patil%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22A.%22%2C%22lastName%22%3A%22Alegre%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Y.%22%2C%22lastName%22%3A%22Freund%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22F.%22%2C%22lastName%22%3A%22Grondin%22%7D%5D%2C%22abstractNote%22%3A%22We%20present%20a%20system%20for%20localizing%20sound%20sources%20in%20a%20room%20with%20several%20ad%20hoc%20microphone%20arrays.%20Each%20circular%20array%20performs%20direction%20of%20arrival%20%28DOA%29%20estimation%20independently%20using%20commercial%20software.%20The%20DOAs%20are%20fed%20to%20a%20fusion%20center%2C%20concatenated%2C%20and%20used%20to%20perform%20the%20localization%20based%20on%20two%20proposed%20methods%2C%20which%20require%20only%20a%20few%20labeled%20source%20locations%20%28anchor%20points%29%20for%20training.%20The%20first%20proposed%20method%20is%20based%20on%20principal%20component%20analysis%20%28PCA%29%20of%20the%20observed%20DOA%20and%20does%20not%20require%20any%20knowledge%20of%20anchor%20points.%20The%20array%20cluster%20can%20then%20perform%20localization%20on%20a%20manifold%20defined%20by%20the%20PCA%20of%20concatenated%20DOAs%20over%20time.%20The%20second%20proposed%20method%20performs%20localization%20using%20an%20affine%20transformation%20between%20the%20DOA%20vectors%20and%20the%20room%20manifold.%20The%20PCA%20has%20fewer%20requirements%20on%20the%20training%20sequence%2C%20but%20is%20less%20robust%20to%20missing%20DOAs%20from%20one%20of%20the%20arrays.%20The%20methods%20are%20demonstrated%20with%20five%20IoT%208-microphone%20circular%20arrays%2C%20placed%20at%20unspecified%20fixed%20locations%20in%20an%20office.%20Both%20the%20PCA%20and%20the%20affine%20method%20can%20easily%20map%20out%20a%20rectangle%20based%20on%20a%20few%20anchor%20points%20with%20similar%20accuracy.%20The%20proposed%20methods%20provide%20a%20step%20toward%20monitoring%20activities%20in%20a%20smart%20home%20and%20require%20little%20installation%20effort%20as%20the%20array%20locations%20are%20not%20needed.%22%2C%22date%22%3A%222022%5C%2F03%22%2C%22language%22%3A%22%22%2C%22DOI%22%3A%2210.1109%5C%2Fjiot.2021.3103523%22%2C%22ISSN%22%3A%222327-4662%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222022-07-13T16%3A48%3A13Z%22%7D%7D%2C%7B%22key%22%3A%22YTLQQT5R%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Zhou%20et%20al.%22%2C%22parsedDate%22%3A%222022-01-16%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EZhou%2C%20Z.%2C%20Bianco%2C%20M.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20%26amp%3B%20Olsen%2C%20K.%20%282022%29.%20High-Resolution%20Imaging%20of%20Complex%20Shallow%20Fault%20Zones%20Along%20the%20July%202019%20Ridgecrest%20Ruptures.%20%3Ci%3EGeophysical%20Research%20Letters%3C%5C%2Fi%3E%2C%20%3Ci%3E49%3C%5C%2Fi%3E%281%29.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2FARTN%20e2021GL095024%2010.1029%5C%2F2021GL095024%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2FARTN%20e2021GL095024%2010.1029%5C%2F2021GL095024%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22High-Resolution%20Imaging%20of%20Complex%20Shallow%20Fault%20Zones%20Along%20the%20July%202019%20Ridgecrest%20Ruptures%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Z.%22%2C%22lastName%22%3A%22Zhou%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22M.%22%2C%22lastName%22%3A%22Bianco%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22K.%22%2C%22lastName%22%3A%22Olsen%22%7D%5D%2C%22abstractNote%22%3A%22We%20perform%20ambient%20noise%20tomography%20using%20data%20recorded%20on%20342%20seismographs%20within%20a%2050%20x%2050%20km%20area%20inside%20which%20the%20July%202019%20M7.1%20and%20M6.4%20Ridgecrest%20earthquakes%20occurred.%20We%20used%20the%20locally%20sparse%20tomography%20%28LST%29%20method%2C%20an%20unsupervised%20machine%20learning%20approach%20that%20learns%20to%20represent%20small-scale%20geophysical%20structures%20using%20only%20data%20from%20the%20immediate%20study.%20The%20Rayleigh%20group%20speed%20obtained%20from%20LST%20better%20predicts%20travel%20times%20than%20conventional%20regularized%20least%20squares%20inversion.%20The%203D%20shear%20velocity%20model%20of%20the%20area%20obtained%20from%20the%20surface%20wave%20dispersion%20maps%20reveals%20a%20highly%20heterogeneous%20low-velocity%20zone%20%28with%20the%20primary%20velocity%20reduction%20in%20the%20upper%202-3%20km%29%20around%20the%20causative%20faults%20for%20the%20M7.1%20and%20M6.4%20events%2C%20with%20a%2040%25%20reduction%20of%20the%20shear%20wave%20velocity.%20Further%2C%20correlation%20of%20other%20imaged%20LVZs%20in%20the%20model%20area%20with%20parts%20of%20the%20Little%20Lake%20Fault%20System%20without%20recent%20activity%20may%20indicate%20long-lasting%20damage%20zones.%22%2C%22date%22%3A%22Jan%2016%202022%22%2C%22language%22%3A%22English%22%2C%22DOI%22%3A%22ARTN%20e2021GL095024%2010.1029%5C%2F2021GL095024%22%2C%22ISSN%22%3A%220094-8276%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%2289L3J9FA%22%5D%2C%22dateModified%22%3A%222024-05-03T16%3A55%3A54Z%22%7D%7D%2C%7B%22key%22%3A%222XGR4FWF%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Whiteaker%20and%20Gerstoft%22%2C%22parsedDate%22%3A%222022%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EWhiteaker%2C%20B.%2C%20%26amp%3B%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%20%282022%29.%20Reducing%20echo%20state%20network%20size%20with%20controllability%20matrices.%20%3Ci%3EChaos%3A%20An%20Interdisciplinary%20Journal%20of%20Nonlinear%20Science%3C%5C%2Fi%3E%2C%20%3Ci%3E32%3C%5C%2Fi%3E%287%29%2C%20073116.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1063%5C%2F5.0071926%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1063%5C%2F5.0071926%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Reducing%20echo%20state%20network%20size%20with%20controllability%20matrices%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Brian%22%2C%22lastName%22%3A%22Whiteaker%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%5D%2C%22abstractNote%22%3A%22Echo%20state%20networks%20are%20a%20fast%20training%20variant%20of%20recurrent%20neural%20networks%20excelling%20at%20approximating%20nonlinear%20dynamical%20systems%20and%20time%20series%20prediction.%20These%20machine%20learning%20models%20act%20as%20nonlinear%20fading%20memory%20filters.%20While%20these%20models%20benefit%20from%20quick%20training%20and%20low%20complexity%2C%20computation%20demands%20from%20a%20large%20reservoir%20matrix%20are%20a%20bottleneck.%20Using%20control%20theory%2C%20a%20reduced%20size%20replacement%20reservoir%20matrix%20is%20found.%20Starting%20from%20a%20large%2C%20task-effective%20reservoir%20matrix%2C%20we%20form%20a%20controllability%20matrix%20whose%20rank%20indicates%20the%20active%20sub-manifold%20and%20candidate%20replacement%20reservoir%20size.%20Resulting%20time%20speed-ups%20and%20reduced%20memory%20usage%20come%20with%20minimal%20error%20increase%20to%20chaotic%20climate%20reconstruction%20or%20short%20term%20prediction.%20Experiments%20are%20performed%20on%20simple%20time%20series%20signals%20and%20the%20Lorenz-1963%20and%20Mackey%5Cu2013Glass%20complex%20chaotic%20signals.%20Observing%20low%20error%20models%20shows%20variation%20of%20active%20rank%20and%20memory%20along%20a%20sequence%20of%20predictions.%22%2C%22date%22%3A%2207%5C%2F2022%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1063%5C%2F5.0071926%22%2C%22ISSN%22%3A%221054-1500%2C%201089-7682%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Faip.scitation.org%5C%2Fdoi%5C%2F10.1063%5C%2F5.0071926%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222023-01-09T23%3A44%3A43Z%22%7D%7D%2C%7B%22key%22%3A%22ZAZ8TKI5%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Hahmann%20et%20al.%22%2C%22parsedDate%22%3A%222022%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EHahmann%2C%20M.%2C%20Fernandez-Grande%2C%20E.%2C%20Gunawan%2C%20H.%2C%20%26amp%3B%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%20%282022%29.%20Sound%20source%20localization%20using%20multiple%20%3Ci%3Ead%20hoc%3C%5C%2Fi%3E%20distributed%20microphone%20arrays.%20%3Ci%3EJASA%20Express%20Letters%3C%5C%2Fi%3E%2C%20%3Ci%3E2%3C%5C%2Fi%3E%287%29%2C%20074801.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0011811%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0011811%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Sound%20source%20localization%20using%20multiple%20%3Ci%3Ead%20hoc%3C%5C%2Fi%3E%20distributed%20microphone%20arrays%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Manuel%22%2C%22lastName%22%3A%22Hahmann%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Efren%22%2C%22lastName%22%3A%22Fernandez-Grande%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Henrry%22%2C%22lastName%22%3A%22Gunawan%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%5D%2C%22abstractNote%22%3A%22Sound%20source%20localization%20is%20crucial%20for%20communication%20and%20sound%20scene%20analysis.%20This%20study%20uses%20direction-of-arrival%20estimates%20of%20multiple%20ad%20hoc%20distributed%20microphone%20arrays%20to%20localize%20sound%20sources%20in%20a%20room.%20An%20affine%20mapping%20between%20the%20independent%20array%20estimates%20and%20the%20source%20coordinates%20is%20derived%20from%20a%20set%20of%20calibration%20points.%20Experiments%20show%20that%20the%20affine%20model%20is%20sufficient%20to%20locate%20a%20source%20and%20can%20be%20calibrated%20to%20physical%20dimensions.%20A%20projection%20of%20the%20local%20array%20estimates%20increases%20localization%20accuracy%2C%20particularly%20further%20away%20from%20the%20calibrated%20region.%20Localization%20tests%20in%20three%20dimensions%20compare%20the%20affine%20approach%20to%20a%20nonlinear%20neural%20network.%22%2C%22date%22%3A%2207%5C%2F2022%22%2C%22language%22%3A%22en%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0011811%22%2C%22ISSN%22%3A%222691-1191%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fasa.scitation.org%5C%2Fdoi%5C%2F10.1121%5C%2F10.0011811%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222022-12-02T18%3A04%3A38Z%22%7D%7D%2C%7B%22key%22%3A%2255TJ8UKJ%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Herchig%20et%20al.%22%2C%22parsedDate%22%3A%222022%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EHerchig%2C%20R.%2C%20Palermo%2C%20N.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20Daniel%2C%20T.%2C%20%26amp%3B%20Sternlicht%2C%20D.%20%282022%29.%20Comparing%20the%20Performance%20of%20Convolutional%20Neural%20Networks%20Trained%20to%20Localize%20Underwater%20Sound%20Sources.%20%3Ci%3E2022%20Oceans%20Hampton%20Roads%3C%5C%2Fi%3E.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2FOceans47191.2022.9977120%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2FOceans47191.2022.9977120%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Comparing%20the%20Performance%20of%20Convolutional%20Neural%20Networks%20Trained%20to%20Localize%20Underwater%20Sound%20Sources%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22R.%22%2C%22lastName%22%3A%22Herchig%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22N.%22%2C%22lastName%22%3A%22Palermo%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22T.%22%2C%22lastName%22%3A%22Daniel%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22D.%22%2C%22lastName%22%3A%22Sternlicht%22%7D%5D%2C%22abstractNote%22%3A%22Localizing%20underwater%20sound%20sources%20in%20the%20highly-variable%20ocean%20environment%20is%20challenging.%20The%20performance%20of%20conventional%20methods%20for%20localizing%20sound%20sources%2C%20such%20as%20matched-field%20processing%2C%20depends%20on%20accurate%20approximations%20of%20the%20environment%20while%20localization%20methods%20using%20neural%20networks%20typically%20generalize%20better%2C%20even%20when%20information%20about%20the%20environment%20is%20sparse.%20This%20paper%20directly%20compares%20the%20performance%20of%20convolutional%20neural%20networks%20%28CNNs%29%20trained%20to%20localize%20sound%20sources%20in%20simulation%20from%20real-valued%20%28magnitude%20only%29%20versus%20complex-valued%20pressure%20data.%20We%20also%20compared%20the%20performance%20of%20these%20two%20types%20of%20networks%20on%20data%20measured%20from%20single%20or%20multiple%20receivers%20in%20different%20spatial%20arrangements%20in%20the%20simulated%20water%20column.%20We%20also%20measured%20the%20CNNs%27%20performance%20loss%20as%20the%20number%20of%20sound%20speed%20profiles%20%28SSPs%29%20used%20to%20generate%20the%20simulated%20field%20replicas%20was%20increased.%20We%20found%20that%20when%20trained%20on%20data%20from%20only%20one%20SSP%2C%20the%20complex-valued%20network%20predicted%20the%20range%20and%20depth%20of%20the%20source%20more%20accurately%20than%20the%20magnitude-only%20network.%20However%2C%20as%20the%20number%20of%20receivers%20increased%2C%20this%20accuracy%20margin%20between%20the%20two%20network%20types%20decreased.%20All%20CNNs%20lost%20accuracy%20as%20we%20increased%20the%20number%20of%20SSPs%2C%20as%20predicted%2C%20but%20the%20magnitude-only%20CNNs%20still%20maintained%20relatively%20high%20accuracy%20for%20depth%20predictions%20even%20after%20training%20on%20data%20from%20all%2015%20different%20SSPs%2C%20exceeding%20the%20accuracy%20of%20the%20complex-valued%20networks.%20For%20range%20inferences%2C%20the%20complex-valued%20networks%20performed%20better%20than%20the%20real-valued%20networks%2C%20but%20when%20trained%20on%20more%20SSPs%2C%20the%20performance%20of%20the%20complex-valued%20networks%20decreased.%22%2C%22date%22%3A%222022%22%2C%22language%22%3A%22English%22%2C%22DOI%22%3A%2210.1109%5C%2FOceans47191.2022.9977120%22%2C%22ISSN%22%3A%220197-7385%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%2289L3J9FA%22%5D%2C%22dateModified%22%3A%222024-05-03T16%3A55%3A54Z%22%7D%7D%2C%7B%22key%22%3A%22FCRFJJRS%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Bell%20et%20al.%22%2C%22parsedDate%22%3A%222022%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EBell%2C%20R.%2C%20Harris%2C%20F.%20J.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20%26amp%3B%20Bharadia%2C%20D.%20%282022%29.%20High%20Resolution%20Spectral%20Analysis%20and%20Signal%20Segregation%20Using%20the%20Polyphase%20Channelizer.%20%3Ci%3E2022%2056th%20Asilomar%20Conference%20on%20Signals%2C%20Systems%2C%20and%20Computers%3C%5C%2Fi%3E%2C%20519%26%23x2013%3B526.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2FIeeeconf56349.2022.10051908%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2FIeeeconf56349.2022.10051908%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22High%20Resolution%20Spectral%20Analysis%20and%20Signal%20Segregation%20Using%20the%20Polyphase%20Channelizer%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22R.%22%2C%22lastName%22%3A%22Bell%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22F.%20J.%22%2C%22lastName%22%3A%22Harris%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22D.%22%2C%22lastName%22%3A%22Bharadia%22%7D%5D%2C%22abstractNote%22%3A%22The%20windowed%20fast%20Fourier%20transform%20%28FFT%29%20is%20the%20ubiquitous%20tool%20for%20spectral%20analysis%20because%20it%20is%20fast%20and%20easy%20to%20use.%20In%20this%20paper%20we%20will%20show%20how%20the%20polyphase%20channelizer%20can%20replace%20the%20FFT%20for%20spectral%20analysis%20and%20produce%20estimates%20of%20the%20power%20spectral%20density%20%28PSD%29%20and%20the%20shorttime%20Fourier%20transform%20%28STFT%29%20spectrogram.%20We%20refer%20to%20the%20polyphase%20channelizer%20equivalents%20of%20these%20estimates%20respectively%20as%20the%20channelizer%20power%20spectral%20density%20%28cPSD%29%20and%20the%20channo-gram.%20It%20will%20be%20shown%20that%20the%20spectral%20resolution%20of%20the%20cPSD%20and%20channogram%20improves%20upon%20the%20well-established%20FFT-based%20estimators%20for%20a%20given%20FFT%20length%20while%20introducing%20minimal%20processing%20overhead.%20We%20will%20then%20discuss%20Parseval%27s%20theorem%20equivalent%20for%20the%20polyphase%20channelizer.%20Finally%2C%20we%20will%20show%20how%20the%20polyphase%20channelizer%20leads%20to%20straightforward%20signal%20segregation%20and%20inversion%20back%20to%20time%20domain%20when%20signals%20of%20random%20and%20unknown%20center%20frequencies%20and%20bandwidth%20are%20encountered.%20This%20is%20an%20important%20enabler%20for%20real-time%20radio%20frequency%20%28RF%29%20signal%20classification%20using%20machine%20learning%20algorithms.%22%2C%22date%22%3A%222022%22%2C%22language%22%3A%22English%22%2C%22DOI%22%3A%2210.1109%5C%2FIeeeconf56349.2022.10051908%22%2C%22ISSN%22%3A%221058-6393%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%2289L3J9FA%22%5D%2C%22dateModified%22%3A%222024-05-03T16%3A55%3A54Z%22%7D%7D%2C%7B%22key%22%3A%22WEKDPBL8%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Park%20et%20al.%22%2C%22parsedDate%22%3A%222022%22%2C%22numChildren%22%3A0%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EPark%2C%20Y.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20%26amp%3B%20Lee%2C%20J.-H.%20%282022%29.%20Difference-Frequency%20MUSIC%20for%20DOAs.%20%3Ci%3EIEEE%20Signal%20Processing%20Letters%3C%5C%2Fi%3E%2C%20%3Ci%3E29%3C%5C%2Fi%3E%2C%202612%26%23x2013%3B2616.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2FLSP.2022.3230365%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2FLSP.2022.3230365%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Difference-Frequency%20MUSIC%20for%20DOAs%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Yongsung%22%2C%22lastName%22%3A%22Park%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Jeung-Hoon%22%2C%22lastName%22%3A%22Lee%22%7D%5D%2C%22abstractNote%22%3A%22%22%2C%22date%22%3A%222022%22%2C%22language%22%3A%22%22%2C%22DOI%22%3A%2210.1109%5C%2FLSP.2022.3230365%22%2C%22ISSN%22%3A%221070-9908%2C%201558-2361%22%2C%22url%22%3A%22https%3A%5C%2F%5C%2Fieeexplore.ieee.org%5C%2Fdocument%5C%2F9992076%5C%2F%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%22A6U9IMQY%22%5D%2C%22dateModified%22%3A%222023-02-17T16%3A35%3A27Z%22%7D%7D%2C%7B%22key%22%3A%22T8TUD799%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Groll%20et%20al.%22%2C%22parsedDate%22%3A%222022%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EGroll%2C%20H.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20Hofer%2C%20M.%2C%20Blumenstein%2C%20J.%2C%20Zemen%2C%20T.%2C%20%26amp%3B%20Mecklenbrauker%2C%20C.%20F.%20%282022%29.%20Scatterer%20Identification%20by%20Atomic%20Norm%20Minimization%20in%20Vehicular%20mm-Wave%20Propagation%20Channels.%20%3Ci%3EIeee%20Access%3C%5C%2Fi%3E%2C%20%3Ci%3E10%3C%5C%2Fi%3E%2C%20102334%26%23x2013%3B102354.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2Faccess.2022.3205616%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2Faccess.2022.3205616%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Scatterer%20Identification%20by%20Atomic%20Norm%20Minimization%20in%20Vehicular%20mm-Wave%20Propagation%20Channels%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22H.%22%2C%22lastName%22%3A%22Groll%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22M.%22%2C%22lastName%22%3A%22Hofer%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22J.%22%2C%22lastName%22%3A%22Blumenstein%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22T.%22%2C%22lastName%22%3A%22Zemen%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22C.%20F.%22%2C%22lastName%22%3A%22Mecklenbrauker%22%7D%5D%2C%22abstractNote%22%3A%22Sparse%20scatterer%20identification%20with%20atomic%20norm%20minimization%20%28ANM%29%20techniques%20in%20the%20delay-Doppler%20domain%20is%20investigated%20for%20a%20vehicle-to-infrastructure%20millimeter%20wave%20propagation%20channel.%20First%2C%20a%20two-dimensional%20ANM%20is%20formulated%20for%20jointly%20estimating%20the%20time-delays%20and%20Doppler%20frequencies%20associated%20with%20individual%20multipath%20components%20%28MPCs%29%20from%20short-time%20Fourier%20transformed%20measurements.%20The%20two-dimensional%20ANM%20is%20formulated%20as%20a%20semi-definite%20program%20and%20promotes%20sparsity%20in%20the%20delay-Doppler%20domain.%20The%20numerical%20complexity%20of%20the%20two-dimensional%20ANM%20limits%20the%20problem%20size%20which%20results%20in%20processing%20limitations%20on%20the%20time-frequency%20sample%20matrix%20size.%20Subsequently%2C%20a%20decoupled%20form%20of%20ANM%20is%20used%20together%20with%20a%20matrix%20pencil%2C%20allowing%20a%20larger%20sample%20matrix%20size.%20Simulations%20show%20that%20spatial%20clusters%20of%20a%20point-scatterers%20with%20small%20cluster%20spread%20are%20suitable%20to%20model%20specular%20reflection%20which%20result%20in%20significant%20MPCs%20and%20the%20successful%20extraction%20of%20their%20delay-Doppler%20parameters.%20The%20decoupled%20ANM%20is%20applied%20to%20vehicle-to-infrastructure%20channel%20sounder%20measurements%20in%20a%20sub-urban%20street%20in%20Vienna%20at%2062%20GHz.%20The%20obtained%20results%20show%20that%20the%20decoupled%20ANM%20successfully%20extracts%20the%20delay-Doppler%20parameters%20in%20high%20resolution%20for%20the%20channel%27s%20significant%20MPCs.%22%2C%22date%22%3A%222022%22%2C%22language%22%3A%22English%22%2C%22DOI%22%3A%2210.1109%5C%2Faccess.2022.3205616%22%2C%22ISSN%22%3A%222169-3536%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222022-11-21T21%3A34%3A59Z%22%7D%7D%2C%7B%22key%22%3A%22Z3TU733R%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Lai%20et%20al.%22%2C%22parsedDate%22%3A%222021-12%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3ELai%2C%20P.%2C%20Amirkulova%2C%20F.%2C%20%26amp%3B%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%20%282021%29.%20Conditional%20Wasserstein%20generative%20adversarial%20networks%20applied%20to%20acoustic%20metamaterial%20design.%20%3Ci%3EJournal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E150%3C%5C%2Fi%3E%286%29%2C%204362%26%23x2013%3B4374.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0008929%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0008929%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Conditional%20Wasserstein%20generative%20adversarial%20networks%20applied%20to%20acoustic%20metamaterial%20design%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Lai%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22F.%22%2C%22lastName%22%3A%22Amirkulova%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%5D%2C%22abstractNote%22%3A%22This%20work%20presents%20a%20method%20for%20the%20reduction%20of%20the%20total%20scattering%20cross%20section%20%28TSCS%29%20for%20a%20planar%20configuration%20of%20cylinders%20by%20means%20of%20generative%20modeling%20and%20deep%20learning.%20Currently%2C%20the%20minimization%20of%20TSCS%20requires%20repeated%20forward%20modelling%20at%20considerable%20computer%20resources%2C%20whereas%20deep%20learning%20can%20do%20this%20more%20efficiently.%20The%20conditional%20Wasserstein%20generative%20adversarial%20networks%20%28cWGANs%29%20model%20is%20proposed%20for%20minimization%20of%20TSCS%20in%20two%20dimensions%20by%20combining%20Wasserstein%20generative%20adversarial%20networks%20with%20convolutional%20neural%20networks%20to%20simulate%20TSCS%20of%20configuration%20of%20rigid%20scatterers.%20The%20proposed%20cWGAN%20model%20is%20enhanced%20by%20adding%20to%20it%20a%20coordinate%20convolution%20%28CoordConv%29%20layer.%20For%20a%20given%20number%20of%20cylinders%2C%20the%20cWGAN%20model%20generates%20images%20of%202D%20configurations%20of%20cylinders%20that%20minimize%20the%20TSCS.%20The%20proposed%20generative%20model%20is%20illustrated%20with%20examples%20for%20planar%20uniform%20configurations%20of%20rigid%20cylinders.%22%2C%22date%22%3A%222021%5C%2F12%22%2C%22language%22%3A%22English%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0008929%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222022-07-13T16%3A48%3A15Z%22%7D%7D%2C%7B%22key%22%3A%22Y499TNF4%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Liu%20et%20al.%22%2C%22parsedDate%22%3A%222021-10%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3ELiu%2C%20R.%20X.%2C%20Bianco%2C%20M.%20J.%2C%20%26amp%3B%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%20%282021%29.%20Automated%20partial%20differential%20equation%20identification.%20%3Ci%3EJournal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E150%3C%5C%2Fi%3E%284%29%2C%202364%26%23x2013%3B2374.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0006444%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0006444%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Automated%20partial%20differential%20equation%20identification%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22R.%20X.%22%2C%22lastName%22%3A%22Liu%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22M.%20J.%22%2C%22lastName%22%3A%22Bianco%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%5D%2C%22abstractNote%22%3A%22%3C%5C%2Fp%3E%20Inspired%20by%20recent%20developments%20in%20data-driven%20methods%20for%20partial%20differential%20equation%20%28PDE%29%20estimation%2C%20we%20usesparse%20modeling%20techniques%20to%20automatically%20estimate%20PDEs%20from%20data.%20A%20dictionary%20consisting%20of%20hypothetical%20PDEterms%20is%20constructed%20using%20numerical%20differentiation.%20Given%20data%2C%20PDE%20terms%20are%20selected%20assuming%20a%20parsimoniousrepresentation%2C%20which%20is%20enforced%20using%20a%20sparsity%20constraint.%20Unlike%20previous%20PDE%20identification%20schemes%2C%20we%20makeno%20assumptions%20about%20which%20PDE%20terms%20are%20responsible%20for%20a%20given%20field.%20The%20approach%20is%20demonstrated%20on%20syntheticand%20real%20video%20data%2C%20with%20physical%20phenomena%20governed%20by%20wave%2C%20Burgers%2C%20and%20Helmholtz%20equations.%20Codes%20areavailable%20athttps%3A%5C%2F%5C%2Fgithub.com%5C%2FNoiseLab-RLiu%5C%2FAutomate-PDE-identification.%20%28C%29%202021%20Acoustical%20Society%20of%20America%3Cbr%3E%3C%5C%2Fp%3E%22%2C%22date%22%3A%222021%5C%2F10%22%2C%22language%22%3A%22English%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0006444%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222022-07-13T16%3A48%3A13Z%22%7D%7D%2C%7B%22key%22%3A%22IRL8PI2H%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Aster%20et%20al.%22%2C%22parsedDate%22%3A%222021-09%22%2C%22numChildren%22%3A4%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EAster%2C%20R.%20C.%2C%20Lipovsky%2C%20B.%20P.%2C%20Cole%2C%20H.%20M.%2C%20Bromirski%2C%20P.%20D.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20Nyblade%2C%20A.%2C%20Wiens%2C%20D.%20A.%2C%20%26amp%3B%20Stephen%2C%20R.%20%282021%29.%20Swell-triggered%20seismicity%20at%20the%20near-front%20damage%20zone%20of%20the%20Ross%20Ice%20Shelf.%20%3Ci%3ESeismological%20Research%20Letters%3C%5C%2Fi%3E%2C%20%3Ci%3E92%3C%5C%2Fi%3E%285%29%2C%202768%26%23x2013%3B2792.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1785%5C%2F0220200478%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1785%5C%2F0220200478%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Swell-triggered%20seismicity%20at%20the%20near-front%20damage%20zone%20of%20the%20Ross%20Ice%20Shelf%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22R.%20C.%22%2C%22lastName%22%3A%22Aster%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22B.%20P.%22%2C%22lastName%22%3A%22Lipovsky%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22H.%20M.%22%2C%22lastName%22%3A%22Cole%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%20D.%22%2C%22lastName%22%3A%22Bromirski%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22A.%22%2C%22lastName%22%3A%22Nyblade%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22D.%20A.%22%2C%22lastName%22%3A%22Wiens%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22R.%22%2C%22lastName%22%3A%22Stephen%22%7D%5D%2C%22abstractNote%22%3A%22Ocean%20swell%20interacting%20with%20Antarctic%20ice%20shelves%20produces%20sustained%20%28approximately%2C%202%20x%20106%20cycles%20per%20year%29%20gravity-elastic%20perturbations%20with%20deformation%20amplitudes%20near%20the%20ice%20front%20as%20large%20as%20tens%20to%20hundreds%20of%20nanostrain.%20This%20process%20is%20the%20most%20energetically%20excited%20during%20the%20austral%20summer%2C%20when%20sea%20ice-induced%20swell%20attenuation%20is%20at%20a%20minimum.%20A%202014-2017%20deployment%20of%20broadband%20seismographs%20on%20the%20Ross%20Ice%20shelf%2C%20which%20included%20three%20stations%20sited%2C%20approximately%2C%202%20km%20from%20the%20ice%20front%2C%20reveals%20prolific%20swell-associated%20triggering%20of%20discrete%20near-ice-front%20%28magnitude%20greater%20than%20or%20similar%20to%200%29%20seismic%20subevents%2C%20for%20which%20we%20identify%20three%20generic%20types.%20During%20some%20strong%20swell%20episodes%2C%20subevent%20timing%20becomes%20sufficiently%20phase-locked%20with%20swell%20excitation%2C%20to%20create%20prominent%20harmonic%20features%20in%20spectra%20calculated%20across%20sufficiently%20lengthy%20time%20windows%20via%20a%20Dirac%20comb%20effect%2C%20for%20which%20we%20articulate%20a%20theoretical%20development%20for%20randomized%20interevent%20times.%20These%20events%20are%20observable%20at%20near-front%20stations%2C%20have%20dominant%20frequency%20content%20between%200.5%20and%2020%20Hz%2C%20and%2C%20in%20many%20cases%2C%20show%20highly%20repetitive%20waveforms.%20Matched%20filtering%20detection%20and%20analysis%20shows%20that%20events%20occur%20at%20a%20low-background%20rate%20during%20all%20swell%20states%2C%20but%20become%20particularly%20strongly%20excited%20during%20large%20amplitude%20swell%20at%20rates%20of%20up%20to%20many%20thousands%20per%20day.%20The%20superimposed%20elastic%20energy%20from%20swell-triggered%20sources%20illuminates%20the%20shelf%20interior%20as%20extensional%20%28elastic%20plate%29%20Lamb%20waves%20that%20are%20observable%20more%20than%20100%20km%20from%20the%20ice%20edge.%20Seismic%20swarms%20show%20threshold%20excitation%20and%20hysteresis%20with%20respect%20to%20rising%20and%20falling%20swell%20excitation.%20This%20behavior%20is%20consistent%20with%20repeated%20seismogenic%20fracture%20excitation%20and%20growth%20within%20a%20near-ice-front%20damage%20zone%2C%20encompassing%20fracture%20features%20seen%20in%20satellite%20imagery.%20A%20much%20smaller%20population%20of%20distinctly%20larger%20near-front%20seismic%20events%2C%20previously%20noted%20to%20be%20weakly%20associated%20with%20extended%20periods%20of%20swell%20perturbation%2C%20likely%20indicate%20calving%20or%20other%20larger-scale%20ice%20failures%20near%20the%20shelf%20front.%22%2C%22date%22%3A%222021%5C%2F09%22%2C%22language%22%3A%22English%22%2C%22DOI%22%3A%2210.1785%5C%2F0220200478%22%2C%22ISSN%22%3A%220895-0695%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22D2UAFEX5%22%2C%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222022-08-05T16%3A09%3A19Z%22%7D%7D%2C%7B%22key%22%3A%22GT6VZQGR%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Jenkins%20et%20al.%22%2C%22parsedDate%22%3A%222021-09%22%2C%22numChildren%22%3A4%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EJenkins%2C%20W.%20F.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20Bianco%2C%20M.%20J.%2C%20%26amp%3B%20Bromirski%2C%20P.%20D.%20%282021%29.%20Unsupervised%20deep%20clustering%20of%20seismic%20data%3A%20Monitoring%20the%20Ross%20Ice%20Shelf%2C%20Antarctica.%20%3Ci%3EJournal%20of%20Geophysical%20Research-Solid%20Earth%3C%5C%2Fi%3E%2C%20%3Ci%3E126%3C%5C%2Fi%3E%289%29%2C%2026.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1029%5C%2F2021jb021716%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1029%5C%2F2021jb021716%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Unsupervised%20deep%20clustering%20of%20seismic%20data%3A%20Monitoring%20the%20Ross%20Ice%20Shelf%2C%20Antarctica%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22W.%20F.%22%2C%22lastName%22%3A%22Jenkins%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22M.%20J.%22%2C%22lastName%22%3A%22Bianco%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%20D.%22%2C%22lastName%22%3A%22Bromirski%22%7D%5D%2C%22abstractNote%22%3A%22Advances%20in%20machine%20learning%20%28ML%29%20techniques%20and%20computational%20capacity%20have%20yielded%20state-of-the-art%20methodologies%20for%20processing%2C%20sorting%2C%20and%20analyzing%20large%20seismic%20data%20sets.%20In%20this%20study%2C%20we%20consider%20an%20application%20of%20ML%20for%20automatically%20identifying%20dominant%20types%20of%20impulsive%20seismicity%20contained%20in%20observations%20from%20a%2034-station%20broadband%20seismic%20array%20deployed%20on%20the%20Ross%20Ice%20Shelf%20%28RIS%29%2C%20Antarctica%20from%202014%20to%202017.%20The%20RIS%20seismic%20data%20contain%20signals%20and%20noise%20generated%20by%20many%20glaciological%20processes%20that%20are%20useful%20for%20monitoring%20the%20integrity%20and%20dynamics%20of%20ice%20shelves.%20Deep%20clustering%20was%20employed%20to%20efficiently%20investigate%20these%20signals.%20Deep%20clustering%20automatically%20groups%20signals%20into%20hypothetical%20classes%20without%20the%20need%20for%20manual%20labeling%2C%20allowing%20for%20the%20comparison%20of%20their%20signal%20characteristics%20and%20spatial%20and%20temporal%20distribution%20with%20potential%20source%20mechanisms.%20The%20method%20uses%20spectrograms%20as%20input%20and%20encodes%20their%20salient%20features%20into%20a%20lower-dimensional%20latent%20representation%20using%20an%20autoencoder%2C%20a%20type%20of%20deep%20neural%20network.%20For%20comparison%2C%20two%20clustering%20methods%20are%20applied%20to%20the%20latent%20data%3A%20a%20Gaussian%20mixture%20model%20%28GMM%29%20and%20deep%20embedded%20clustering%20%28DEC%29.%20Eight%20classes%20of%20dominant%20seismic%20signals%20were%20identified%20and%20compared%20with%20environmental%20data%20such%20as%20temperature%2C%20wind%20speed%2C%20tides%2C%20and%20sea%20ice%20concentration.%20The%20greatest%20seismicity%20levels%20occurred%20at%20the%20RIS%20front%20during%20the%202016%20El%20Nino%20summer%2C%20and%20near%20grounding%20zones%20near%20the%20front%20throughout%20the%20deployment.%20We%20demonstrate%20the%20spatial%20and%20temporal%20association%20of%20certain%20classes%20of%20seismicity%20with%20seasonal%20changes%20at%20the%20RIS%20front%2C%20and%20with%20tidally%20driven%20seismicity%20at%20Roosevelt%20Island.%22%2C%22date%22%3A%222021%5C%2F09%22%2C%22language%22%3A%22English%22%2C%22DOI%22%3A%2210.1029%5C%2F2021jb021716%22%2C%22ISSN%22%3A%222169-9313%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22D2UAFEX5%22%2C%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222022-08-02T22%3A21%3A28Z%22%7D%7D%2C%7B%22key%22%3A%22TKRYSTFX%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Li%20et%20al.%22%2C%22parsedDate%22%3A%222021-09%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3ELi%2C%20J.%2C%20Siderius%2C%20M.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20Fan%2C%20J.%2C%20%26amp%3B%20Muzi%2C%20L.%20%282021%29.%20Head-wave%20correlations%20in%20layered%20seabed%3A%20Theory%20and%20modeling.%20%3Ci%3EJASA%20Express%20Letters%3C%5C%2Fi%3E%2C%20%3Ci%3E1%3C%5C%2Fi%3E%289%29%2C%207.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0006390%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0006390%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Head-wave%20correlations%20in%20layered%20seabed%3A%20Theory%20and%20modeling%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22J.%22%2C%22lastName%22%3A%22Li%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22M.%22%2C%22lastName%22%3A%22Siderius%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22J.%22%2C%22lastName%22%3A%22Fan%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22L.%22%2C%22lastName%22%3A%22Muzi%22%7D%5D%2C%22abstractNote%22%3A%22This%20paper%20derives%20travel%20times%20and%20arrival%20angles%20of%20head-wave%20correlations%20from%20ocean%20ambient%20noise%20in%20shallow%20water%20over%20a%20layered%20seabed.%20The%20upcoming%20and%20surface%20reflected%20head-wave%20noise%20signal%20received%20at%20two%20receivers%20from%20the%20same%20interface%20are%20correlated%2C%20and%20their%20travel%20time%20differences%20give%20the%20travel%20times%20of%20the%20head-wave%20correlations.%20The%20arrival%20angle%20of%20head-wave%20correlations%20from%20an%20interface%20depends%20on%20sound%20speeds%20in%20the%20layers%20above%20and%20just%20below.%20The%20predictions%20of%20head-wave%20correlations%20from%20a%20seabed%20with%20two%20layers%20and%20the%20corresponding%20inversion%20results%20are%20verified%20with%20simulations.%20%28C%29%202021%20Author%28s%29.%20All%20article%20content%2C%20except%20where%20otherwise%20noted%2C%20is%20licensed%20under%20a%20Creative%20Commons%20Attribution%20%28CC%20BY%29%20license%20%28http%3A%5C%2F%5C%2Fcreativecommons.org%5C%2Flicenses%5C%2Fby%5C%2F4.0%5C%2F%29.%22%2C%22date%22%3A%222021%5C%2F09%22%2C%22language%22%3A%22English%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0006390%22%2C%22ISSN%22%3A%22%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222022-07-13T16%3A48%3A18Z%22%7D%7D%2C%7B%22key%22%3A%22EC2Z7858%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Shah%20et%20al.%22%2C%22parsedDate%22%3A%222021-07%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EShah%2C%20T.%2C%20Zhuo%2C%20L.%2C%20Lai%2C%20P.%2C%20De%20la%20Rosa-Moreno%2C%20A.%2C%20Amirkulova%2C%20F.%2C%20%26amp%3B%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%20%282021%29.%20Reinforcement%20learning%20applied%20to%20metamaterial%20design.%20%3Ci%3EJournal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E150%3C%5C%2Fi%3E%281%29%2C%20321%26%23x2013%3B338.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0005545%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0005545%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Reinforcement%20learning%20applied%20to%20metamaterial%20design%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Tristan%22%2C%22lastName%22%3A%22Shah%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Linwei%22%2C%22lastName%22%3A%22Zhuo%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Lai%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Amaris%22%2C%22lastName%22%3A%22De%20la%20Rosa-Moreno%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Feruza%22%2C%22lastName%22%3A%22Amirkulova%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Peter%22%2C%22lastName%22%3A%22Gerstoft%22%7D%5D%2C%22abstractNote%22%3A%22This%20paper%20presents%20a%20semi-analytical%20method%20of%20suppressing%20acoustic%20scattering%20using%20reinforcement%20learning%20%28RL%29%20algorithms.%20We%20give%20a%20RL%20agent%20control%20over%20design%20parameters%20of%20a%20planar%20configuration%20of%20cylindrical%20scatterers%20in%20water.%20These%20design%20parameters%20control%20the%20position%20and%20radius%20of%20the%20scatterers.%20As%20these%20cylinders%20encounter%20an%20incident%20acoustic%20wave%2C%20the%20scattering%20pattern%20is%20described%20by%20a%20function%20called%20total%20scattering%20cross%20section%20%28TSCS%29.%20Through%20evaluating%20the%20gradients%20of%20TSCS%20and%20other%20information%20about%20the%20state%20of%20the%20configuration%2C%20the%20RL%20agent%20perturbatively%20adjusts%20design%20parameters%2C%20considering%20multiple%20scattering%20between%20the%20scatterers.%20As%20each%20adjustment%20is%20made%2C%20the%20RL%20agent%20receives%20a%20reward%20negatively%20proportional%20to%20the%20root%20mean%20square%20of%20the%20TSCS%20across%20a%20range%20of%20wavenumbers.%20Through%20maximizing%20its%20reward%20per%20episode%2C%20the%20agent%20discovers%20designs%20with%20low%20scattering.%20Specifically%2C%20the%20double%20deep%20Q-learning%20network%20and%20the%20deep%20deterministic%20policy%20gradient%20algorithms%20are%20employed%20in%20our%20models.%20Designs%20discovered%20by%20the%20RL%20algorithms%20performed%20well%20when%20compared%20to%20a%20state-of-the-art%20optimization%20algorithm%20using%20fmincon.%22%2C%22date%22%3A%222021%5C%2F07%22%2C%22language%22%3A%22%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0005545%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222022-07-13T16%3A48%3A23Z%22%7D%7D%2C%7B%22key%22%3A%228QDUC9YI%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Niu%20et%20al.%22%2C%22parsedDate%22%3A%222021-06%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3ENiu%2C%20H.%20Q.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20Zhang%2C%20R.%20H.%2C%20Li%2C%20Z.%20L.%2C%20Gong%2C%20Z.%20X.%2C%20%26amp%3B%20Wang%2C%20H.%20B.%20%282021%29.%20Mode%20separation%20with%20one%20hydrophone%20in%20shallow%20water%3A%20A%20sparse%20Bayesian%20learning%20approach%20based%20on%20phase%20speed.%20%3Ci%3EJournal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E149%3C%5C%2Fi%3E%286%29%2C%204366%26%23x2013%3B4376.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0005312%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0005312%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Mode%20separation%20with%20one%20hydrophone%20in%20shallow%20water%3A%20A%20sparse%20Bayesian%20learning%20approach%20based%20on%20phase%20speed%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22H.%20Q.%22%2C%22lastName%22%3A%22Niu%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22R.%20H.%22%2C%22lastName%22%3A%22Zhang%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Z.%20L.%22%2C%22lastName%22%3A%22Li%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Z.%20X.%22%2C%22lastName%22%3A%22Gong%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22H.%20B.%22%2C%22lastName%22%3A%22Wang%22%7D%5D%2C%22abstractNote%22%3A%22An%20approach%20of%20broadband%20mode%20separation%20in%20shallow%20water%20is%20proposed%20using%20phase%20speed%20extracted%20from%20one%20hydrophone%20and%20solved%20with%20sparse%20Bayesian%20learning%20%28SBL%29.%20The%20approximate%20modal%20dispersion%20relation%2C%20connecting%20the%20horizontal%20wavenumbers%20%28phase%20velocities%29%20for%20multiple%20frequencies%2C%20is%20used%20to%20build%20the%20dictionary%20matrix%20for%20SBL.%20Given%20a%20multi-frequency%20pressure%20vector%20on%20one%20hydrophone%2C%20SBL%20estimates%20a%20set%20of%20sparse%20coefficients%20for%20a%20large%20number%20of%20atoms%20in%20the%20dictionary.%20With%20the%20estimated%20coefficients%20and%20corresponding%20atoms%2C%20the%20separated%20normal%20modes%20are%20retrieved.%20The%20presented%20method%20can%20be%20used%20for%20impulsive%20or%20known-form%20signals%20in%20a%20shallow-water%20environment%20while%20no%20bottom%20information%20is%20required.%20The%20simulation%20results%20demonstrate%20that%20the%20proposed%20approach%20is%20adapted%20to%20the%20environment%20where%20both%20the%20reflected%20and%20refracted%20modes%20coexist%2C%20whereas%20the%20performance%20of%20the%20time%20warping%20transformation%20degrades%20significantly%20in%20this%20scenario.%22%2C%22date%22%3A%222021%5C%2F06%22%2C%22language%22%3A%22%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0005312%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222022-07-13T16%3A48%3A20Z%22%7D%7D%2C%7B%22key%22%3A%22ZK42Q3FQ%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Wagner%20et%20al.%22%2C%22parsedDate%22%3A%222021-04%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EWagner%2C%20M.%2C%20Park%2C%20Y.%2C%20%26amp%3B%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%20%282021%29.%20Gridless%20DOA%20estimation%20and%20root-MUSIC%20for%20non-uniform%20linear%20arrays.%20%3Ci%3EIeee%20Transactions%20on%20Signal%20Processing%3C%5C%2Fi%3E%2C%20%3Ci%3E69%3C%5C%2Fi%3E%2C%202144%26%23x2013%3B2157.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2Ftsp.2021.3068353%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1109%5C%2Ftsp.2021.3068353%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Gridless%20DOA%20estimation%20and%20root-MUSIC%20for%20non-uniform%20linear%20arrays%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22M.%22%2C%22lastName%22%3A%22Wagner%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Y.%22%2C%22lastName%22%3A%22Park%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%5D%2C%22abstractNote%22%3A%22Gridless%20direction%20of%20arrival%20%28DOA%29%20estimation%20is%20addressed%20for%20a%201-D%20non-uniform%20array%20%28NUA%29%20of%20arbitrary%20geometry.%20Currently%2C%20gridless%20DOA%20estimation%20is%20solved%20via%20convex%20relaxation%2C%20and%20is%20applicable%20only%20to%20uniform%20linear%20arrays%20%28ULA%29.%20The%20ULA%20sample%20covariance%20matrix%20has%20Toeplitz%20structure%2C%20and%20gridless%20DOA%20is%20based%20on%20the%20Vandermonde%20decomposition%20of%20this%20matrix.%20The%20Vandermonde%20decomposition%20breaks%20a%20Toeplitz%20matrix%20into%20its%20harmonic%20components%2C%20from%20which%20the%20DOAs%20are%20estimated.%20First%2C%20we%20present%20the%20irregular%20Toeplitz%20matrix%20and%20irregular%20Vandermonde%20decomposition%20%28IVD%29%2C%20which%20generalizes%20the%20Vandermonde%20decomposition.%20It%20is%20shown%20that%20IVD%20is%20related%20to%20the%20MUSIC%20and%20root-MUSIC%20algorithms.%20Next%2C%20atomic%20norm%20minimization%20%28ANM%29%20for%20gridless%20DOA%20is%20generalized%20to%20NUA%20using%20the%20IVD.%20The%20resulting%20non-convex%20optimization%20is%20solved%20using%20alternating%20projections%20%28AP%29.%20Simulations%20show%20the%20AP%20based%20solution%20for%20NUA%5C%2FULA%20has%20similar%20accuracy%20as%20convex%20relaxation%20of%20gridless%20DOA%20for%20ULA.%22%2C%22date%22%3A%222021%5C%2F04%22%2C%22language%22%3A%22%22%2C%22DOI%22%3A%2210.1109%5C%2Ftsp.2021.3068353%22%2C%22ISSN%22%3A%221053-587X%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%22A6U9IMQY%22%5D%2C%22dateModified%22%3A%222022-07-13T16%3A48%3A18Z%22%7D%7D%2C%7B%22key%22%3A%22GJ9THN5R%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Ozanich%20et%20al.%22%2C%22parsedDate%22%3A%222021-04%22%2C%22numChildren%22%3A4%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EOzanich%2C%20E.%2C%20Thode%2C%20A.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20Freeman%2C%20L.%20A.%2C%20%26amp%3B%20Freeman%2C%20S.%20%282021%29.%20Deep%20embedded%20clustering%20of%20coral%20reef%20bioacoustics.%20%3Ci%3EJournal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E149%3C%5C%2Fi%3E%284%29%2C%202587%26%23x2013%3B2601.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0004221%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0004221%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Deep%20embedded%20clustering%20of%20coral%20reef%20bioacoustics%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22E.%22%2C%22lastName%22%3A%22Ozanich%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22A.%22%2C%22lastName%22%3A%22Thode%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22L.%20A.%22%2C%22lastName%22%3A%22Freeman%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22S.%22%2C%22lastName%22%3A%22Freeman%22%7D%5D%2C%22abstractNote%22%3A%22Deep%20clustering%20was%20applied%20to%20unlabeled%2C%20automatically%20detected%20signals%20in%20a%20coral%20reef%20soundscape%20to%20distinguish%20fish%20pulse%20calls%20from%20segments%20of%20whale%20song.%20Deep%20embedded%20clustering%20%28DEC%29%20learned%20latent%20features%20and%20formed%20classification%20clusters%20using%20fixed-length%20power%20spectrograms%20of%20the%20signals.%20Handpicked%20spectral%20and%20temporal%20features%20were%20also%20extracted%20and%20clustered%20with%20Gaussian%20mixture%20models%20%28GMM%29%20and%20conventional%20clustering.%20DEC%2C%20GMM%2C%20and%20conventional%20clustering%20were%20tested%20on%20simulated%20datasets%20of%20fish%20pulse%20calls%20%28fish%29%20and%20whale%20song%20units%20%28whale%29%20with%20randomized%20bandwidth%2C%20duration%2C%20and%20SNR.%20Both%20GMM%20and%20DEC%20achieved%20high%20accuracy%20and%20identified%20clusters%20with%20fish%2C%20whale%2C%20and%20overlapping%20fish%20and%20whale%20signals.%20Conventional%20clustering%20methods%20had%20low%20accuracy%20in%20scenarios%20with%20unequal-sized%20clusters%20or%20overlapping%20signals.%20Fish%20and%20whale%20signals%20recorded%20near%20Hawaii%20in%20February-March%202020%20were%20clustered%20with%20DEC%2C%20GMM%2C%20and%20conventional%20clustering.%20DEC%20features%20demonstrated%20the%20highest%20accuracy%20of%2077.5%25%20on%20a%20small%2C%20manually%20labeled%20dataset%20for%20classifying%20signals%20into%20fish%20and%20whale%20clusters.%22%2C%22date%22%3A%222021%5C%2F04%22%2C%22language%22%3A%22%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0004221%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%22WFYZNCMA%22%5D%2C%22dateModified%22%3A%222022-09-19T23%3A02%3A24Z%22%7D%7D%2C%7B%22key%22%3A%22AGDG2WTZ%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Park%20et%20al.%22%2C%22parsedDate%22%3A%222021-03%22%2C%22numChildren%22%3A4%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EPark%2C%20Y.%2C%20Meyer%2C%20F.%2C%20%26amp%3B%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%20%282021%29.%20Sequential%20sparse%20Bayesian%20learning%20for%20time-varying%20direction%20of%20arrival.%20%3Ci%3EJournal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E149%3C%5C%2Fi%3E%283%29%2C%202089%26%23x2013%3B2099.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0003802%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0003802%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Sequential%20sparse%20Bayesian%20learning%20for%20time-varying%20direction%20of%20arrival%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Y.%22%2C%22lastName%22%3A%22Park%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22F.%22%2C%22lastName%22%3A%22Meyer%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%5D%2C%22abstractNote%22%3A%22This%20paper%20presents%20methods%20for%20the%20estimation%20of%20the%20time-varying%20directions%20of%20arrival%20%28DOAs%29%20of%20signals%20emitted%20by%20moving%20sources.%20Following%20the%20sparse%20Bayesian%20learning%20%28SBL%29%20framework%2C%20prior%20information%20of%20unknown%20source%20amplitudes%20is%20modeled%20as%20a%20multi-variate%20Gaussian%20distribution%20with%20zero-mean%20and%20time-varying%20variance%20parameters.%20For%20sequential%20estimation%20of%20the%20unknown%20variance%2C%20we%20present%20two%20sequential%20SBL-based%20methods%20that%20propagate%20statistical%20information%20across%20time%20to%20improve%20DOA%20estimation%20performance.%20The%20first%20method%20heuristically%20calculates%20the%20parameters%20of%20an%20inverse-gamma%20hyperprior%20based%20on%20the%20source%20signal%20estimate%20from%20the%20previous%20time%20step.%20In%20addition%2C%20a%20second%20sequential%20SBL%20method%20is%20proposed%2C%20which%20performs%20a%20prediction%20step%20to%20calculate%20the%20prior%20distribution%20of%20the%20current%20variance%20parameter%20from%20the%20variance%20parameter%20estimated%20at%20the%20previous%20time%20step.%20The%20SBL-based%20sequential%20processing%20provides%20high-resolution%20DOA%20tracking%20capabilities.%20Performance%20improvements%20are%20demonstrated%20by%20using%20simulated%20data%20as%20well%20as%20real%20data%20from%20the%20SWellEx-96%20experiment.%22%2C%22date%22%3A%222021%5C%2F03%22%2C%22language%22%3A%22%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0003802%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%22764DPZJV%22%2C%22A6U9IMQY%22%5D%2C%22dateModified%22%3A%222022-09-22T23%3A33%3A16Z%22%7D%7D%2C%7B%22key%22%3A%22V623LS2B%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Cao%20et%20al.%22%2C%22parsedDate%22%3A%222021-03%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3ECao%2C%20H.%20G.%2C%20Wang%2C%20W.%20B.%2C%20Su%2C%20L.%2C%20Ni%2C%20H.%20Y.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20Ren%2C%20Q.%20Y.%2C%20%26amp%3B%20Ma%2C%20L.%20%282021%29.%20Deep%20transfer%20learning%20for%20underwater%20direction%20of%20arrival%20using%20one%20vector%20sensora%29.%20%3Ci%3EJournal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E149%3C%5C%2Fi%3E%283%29%2C%201699%26%23x2013%3B1711.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0003645%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0003645%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Deep%20transfer%20learning%20for%20underwater%20direction%20of%20arrival%20using%20one%20vector%20sensora%29%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22H.%20G.%22%2C%22lastName%22%3A%22Cao%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22W.%20B.%22%2C%22lastName%22%3A%22Wang%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22L.%22%2C%22lastName%22%3A%22Su%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22H.%20Y.%22%2C%22lastName%22%3A%22Ni%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22Q.%20Y.%22%2C%22lastName%22%3A%22Ren%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22L.%22%2C%22lastName%22%3A%22Ma%22%7D%5D%2C%22abstractNote%22%3A%22A%20deep%20transfer%20learning%20%28DTL%29%20method%20is%20proposed%20for%20the%20direction%20of%20arrival%20%28DOA%29%20estimation%20using%20a%20single-vector%20sensor.%20The%20method%20involves%20training%20of%20a%20convolutional%20neural%20network%20%28CNN%29%20with%20synthetic%20data%20in%20source%20domain%20and%20then%20adapting%20the%20source%20domain%20to%20target%20domain%20with%20available%20at-sea%20data.%20The%20CNN%20is%20fed%20with%20the%20cross-spectrum%20of%20acoustical%20pressure%20and%20particle%20velocity%20during%20the%20training%20process%20to%20learn%20DOAs%20of%20a%20moving%20surface%20ship.%20For%20domain%20adaptation%2C%20first%20convolutional%20layers%20of%20the%20pre-trained%20CNN%20are%20copied%20to%20a%20target%20CNN%2C%20and%20the%20remaining%20layers%20of%20the%20target%20CNN%20are%20randomly%20initialized%20and%20trained%20on%20at-sea%20data.%20Numerical%20tests%20and%20real%20data%20results%20suggest%20that%20the%20DTL%20yields%20more%20reliable%20DOA%20estimates%20than%20a%20conventional%20CNN%2C%20especially%20with%20interfering%20sources.%22%2C%22date%22%3A%222021%5C%2F03%22%2C%22language%22%3A%22%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0003645%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222022-07-13T16%3A48%3A15Z%22%7D%7D%2C%7B%22key%22%3A%224XQ89FK2%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Snover%20et%20al.%22%2C%22parsedDate%22%3A%222021-03%22%2C%22numChildren%22%3A2%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3ESnover%2C%20D.%2C%20Johnson%2C%20C.%20W.%2C%20Bianco%2C%20M.%20J.%2C%20%26amp%3B%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%20%282021%29.%20Deep%20clustering%20to%20identify%20sources%20of%20urban%20seismic%20noise%20in%20Long%20Beach%2C%20California.%20%3Ci%3ESeismological%20Research%20Letters%3C%5C%2Fi%3E%2C%20%3Ci%3E92%3C%5C%2Fi%3E%282%29%2C%201011%26%23x2013%3B1022.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1785%5C%2F0220200164%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1785%5C%2F0220200164%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Deep%20clustering%20to%20identify%20sources%20of%20urban%20seismic%20noise%20in%20Long%20Beach%2C%20California%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22D.%22%2C%22lastName%22%3A%22Snover%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22C.%20W.%22%2C%22lastName%22%3A%22Johnson%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22M.%20J.%22%2C%22lastName%22%3A%22Bianco%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%5D%2C%22abstractNote%22%3A%22Ambient%20seismic%20noise%20consists%20of%20emergent%20and%20impulsive%20signals%20generated%20by%20natural%20and%20anthropogenic%20sources.%20Developing%20techniques%20to%20identify%20specific%20cultural%20noise%20signals%20will%20benefit%20studies%20performing%20seismic%20imaging%20from%20continuous%20records.%20We%20examine%20spectrograms%20of%20urban%20cultural%20noise%20from%20a%20spatially%20dense%20seismic%20array%20located%20in%20Long%20Beach%2C%20California.%20The%20spectral%20features%20of%20the%20waveforms%20are%20used%20to%20develop%20a%20self-supervised%20clustering%20model%20for%20differentiating%20cultural%20noise%20into%20separable%20types%20of%20signals.%20We%20use%20161%20hr%20of%20seismic%20data%20from%205200%20geophones%20that%20contain%20impulsive%20signals%20originating%20from%20human%20activity.%20The%20model%20uses%20convolutional%20autoencoders%2C%20a%20self-supervised%20machine-learning%20technique%2C%20to%20learn%20latent%20features%20from%20spectrograms%20produced%20from%20the%20data.%20The%20latent%20features%20are%20evaluated%20using%20a%20deep%20clustering%20algorithm%20to%20separate%20the%20noise%20signals%20into%20different%20classes.%20We%20evaluate%20the%20separation%20of%20data%20and%20analyze%20the%20classes%20to%20identify%20the%20likely%20sources%20of%20the%20signals%20present%20in%20the%20data.%20To%20interpret%20the%20model%20performance%2C%20we%20examine%20the%20time-frequency%20domain%20features%20of%20the%20signals%20and%20the%20spatiotemporal%20evolution%20observed%20for%20each%20class.%20We%20demonstrate%20that%20clustering%20using%20deep%20autoencoders%20is%20a%20useful%20approach%20to%20characterizing%20seismic%20noise%20and%20identifying%20novel%20signals%20in%20the%20data.%22%2C%22date%22%3A%222021%5C%2F03%22%2C%22language%22%3A%22%22%2C%22DOI%22%3A%2210.1785%5C%2F0220200164%22%2C%22ISSN%22%3A%220895-0695%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222022-07-13T16%3A48%3A15Z%22%7D%7D%2C%7B%22key%22%3A%227T4YN5XH%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Thode%20et%20al.%22%2C%22parsedDate%22%3A%222021-02%22%2C%22numChildren%22%3A4%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EThode%2C%20A.%20M.%2C%20Conrad%2C%20A.%20S.%2C%20Ozanich%2C%20E.%2C%20King%2C%20R.%2C%20Freeman%2C%20S.%20E.%2C%20Freeman%2C%20L.%20A.%2C%20Zgliczynski%2C%20B.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20%26amp%3B%20Kim%2C%20K.%20H.%20%282021%29.%20Automated%20two-dimensional%20localization%20of%20underwater%20acoustic%20transient%20impulses%20using%20vector%20sensor%20image%20processing%20%28vector%20sensor%20localization%29.%20%3Ci%3EJournal%20of%20the%20Acoustical%20Society%20of%20America%3C%5C%2Fi%3E%2C%20%3Ci%3E149%3C%5C%2Fi%3E%282%29%2C%20770%26%23x2013%3B787.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0003382%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1121%5C%2F10.0003382%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Automated%20two-dimensional%20localization%20of%20underwater%20acoustic%20transient%20impulses%20using%20vector%20sensor%20image%20processing%20%28vector%20sensor%20localization%29%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22A.%20M.%22%2C%22lastName%22%3A%22Thode%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22A.%20S.%22%2C%22lastName%22%3A%22Conrad%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22E.%22%2C%22lastName%22%3A%22Ozanich%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22R.%22%2C%22lastName%22%3A%22King%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22S.%20E.%22%2C%22lastName%22%3A%22Freeman%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22L.%20A.%22%2C%22lastName%22%3A%22Freeman%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22B.%22%2C%22lastName%22%3A%22Zgliczynski%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22K.%20H.%22%2C%22lastName%22%3A%22Kim%22%7D%5D%2C%22abstractNote%22%3A%22Detecting%20acoustic%20transients%20by%20signal-to-noise%20ratio%20%28SNR%29%20becomes%20problematic%20in%20nonstationary%20ambient%20noise%20environments%20characteristic%20of%20coral%20reefs.%20An%20alternate%20approach%20presented%20here%20uses%20signal%20directionality%20to%20automatically%20detect%20and%20localize%20transient%20impulsive%20sounds%20collected%20on%20underwater%20vector%20sensors%20spaced%20tens%20of%20meters%20apart.%20The%20procedure%2C%20which%20does%20not%20require%20precise%20time%20synchronization%2C%20first%20constructs%20time-frequency%20representations%20of%20both%20the%20squared%20acoustic%20pressure%20%28spectrogram%29%20and%20dominant%20directionality%20of%20the%20active%20intensity%20%28azigram%29%20on%20each%20sensor.%20Within%20each%20azigram%2C%20sets%20of%20time-frequency%20cells%20associated%20with%20transient%20energy%20arriving%20from%20a%20consistent%20azimuthal%20sector%20are%20identified.%20Binary%20image%20processing%20techniques%20then%20link%20sets%20that%20share%20similar%20duration%20and%20bandwidth%20between%20different%20sensors%2C%20after%20which%20the%20algorithm%20triangulates%20the%20source%20location.%20Unlike%20most%20passive%20acoustic%20detectors%2C%20the%20threshold%20criterion%20for%20this%20algorithm%20is%20bandwidth%20instead%20of%20pressure%20magnitude.%20Data%20collected%20from%20shallow%20coral%20reef%20environments%20demonstrate%20the%20algorithm%27s%20ability%20to%20detect%20SCUBA%20bubble%20plumes%20and%20consistent%20spatial%20distributions%20of%20somniferous%20fish%20activity.%20Analytical%20estimates%20and%20direct%20evaluations%20both%20yield%20false%20transient%20localization%20rates%20from%203%25%20to%206%25%20in%20a%20coral%20reef%20environment.%20The%20SNR%20distribution%20of%20localized%20pulses%20off%20Hawaii%20has%20a%20median%20of%207.7%20dB%20and%20interquartile%20range%20of%207.1%20dB.%22%2C%22date%22%3A%222021%5C%2F02%22%2C%22language%22%3A%22%22%2C%22DOI%22%3A%2210.1121%5C%2F10.0003382%22%2C%22ISSN%22%3A%220001-4966%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22EY7L9SEU%22%2C%22WFYZNCMA%22%5D%2C%22dateModified%22%3A%222022-09-22T23%3A33%3A34Z%22%7D%7D%2C%7B%22key%22%3A%22I6822L5X%22%2C%22library%22%3A%7B%22id%22%3A9129767%7D%2C%22meta%22%3A%7B%22creatorSummary%22%3A%22Baker%20et%20al.%22%2C%22parsedDate%22%3A%222021-02%22%2C%22numChildren%22%3A4%7D%2C%22bib%22%3A%22%3Cdiv%20class%3D%5C%22csl-bib-body%5C%22%20style%3D%5C%22line-height%3A%202%3B%20padding-left%3A%201em%3B%20text-indent%3A-1em%3B%5C%22%3E%5Cn%20%20%3Cdiv%20class%3D%5C%22csl-entry%5C%22%3EBaker%2C%20M.%20G.%2C%20Aster%2C%20R.%20C.%2C%20Wiens%2C%20D.%20A.%2C%20Nyblade%2C%20A.%2C%20Bromirski%2C%20P.%20D.%2C%20%3Cstrong%3EGerstoft%3C%5C%2Fstrong%3E%2C%20P.%2C%20%26amp%3B%20Stephen%2C%20R.%20A.%20%282021%29.%20Teleseismic%20earthquake%20wavefields%20observed%20on%20the%20Ross%20Ice%20Shelf.%20%3Ci%3EJournal%20of%20Glaciology%3C%5C%2Fi%3E%2C%20%3Ci%3E67%3C%5C%2Fi%3E%28261%29%2C%2058%26%23x2013%3B74.%20%3Ca%20class%3D%27zp-DOIURL%27%20href%3D%27https%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1017%5C%2Fjog.2020.83%27%3Ehttps%3A%5C%2F%5C%2Fdoi.org%5C%2F10.1017%5C%2Fjog.2020.83%3C%5C%2Fa%3E%3C%5C%2Fdiv%3E%5Cn%3C%5C%2Fdiv%3E%22%2C%22data%22%3A%7B%22itemType%22%3A%22journalArticle%22%2C%22title%22%3A%22Teleseismic%20earthquake%20wavefields%20observed%20on%20the%20Ross%20Ice%20Shelf%22%2C%22creators%22%3A%5B%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22M.%20G.%22%2C%22lastName%22%3A%22Baker%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22R.%20C.%22%2C%22lastName%22%3A%22Aster%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22D.%20A.%22%2C%22lastName%22%3A%22Wiens%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22A.%22%2C%22lastName%22%3A%22Nyblade%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%20D.%22%2C%22lastName%22%3A%22Bromirski%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22P.%22%2C%22lastName%22%3A%22Gerstoft%22%7D%2C%7B%22creatorType%22%3A%22author%22%2C%22firstName%22%3A%22R.%20A.%22%2C%22lastName%22%3A%22Stephen%22%7D%5D%2C%22abstractNote%22%3A%22Observations%20of%20teleseismic%20earthquakes%20using%20broadband%20seismometers%20on%20the%20Ross%20Ice%20Shelf%20%28RIS%29%20must%20contend%20with%20environmental%20and%20structural%20processes%20that%20do%20not%20exist%20for%20land-sited%20seismometers.%20Important%20considerations%20are%3A%20%281%29%20a%20broadband%2C%20multi-mode%20ambient%20wavefield%20excited%20by%20ocean%20gravity%20wave%20interactions%20with%20the%20ice%20shelf%3B%20%282%29%20body%20wave%20reverberations%20produced%20by%20seismic%20impedance%20contrasts%20at%20the%20ice%5C%2Fwater%20and%20water%5C%2Fseafloor%20interfaces%20and%20%283%29%20decoupling%20of%20the%20solid%20Earth%20horizontal%20wavefield%20by%20the%20sub-shelf%20water%20column.%20We%20analyze%20seasonal%20and%20geographic%20variations%20in%20signal-to-noise%20ratios%20for%20teleseismic%20P-wave%20%280.5-2.0%20s%29%2C%20S-wave%20%2810-15%20s%29%20and%20surface%20wave%20%2813-25%20s%29%20arrivals%20relative%20to%20the%20RIS%20noise%20field.%20We%20use%20ice%20and%20water%20layer%20reverberations%20generated%20by%20teleseismic%20P-waves%20to%20accurately%20estimate%20the%20sub-station%20thicknesses%20of%20these%20layers.%20We%20present%20observations%20consistent%20with%20the%20theoretically%20predicted%20transition%20of%20the%20water%20column%20from%20compressible%20to%20incompressible%20mechanics%2C%20relevant%20for%20vertically%20incident%20solid%20Earth%20waves%20with%20periods%20longer%20than%203%20s.%20Finally%2C%20we%20observe%20symmetric-mode%20Lamb%20waves%20generated%20by%20teleseismic%20S-waves%20incident%20on%20the%20grounding%20zones.%20Despite%20their%20complexity%2C%20we%20conclude%20that%20teleseismic%20coda%20can%20be%20utilized%20for%20passive%20imaging%20of%20sub-shelf%20Earth%20structure%2C%20although%20longer%20deployments%20relative%20to%20conventional%20land-sited%20seismometers%20will%20be%20necessary%20to%20acquire%20adequate%20data.%22%2C%22date%22%3A%222021%5C%2F02%22%2C%22language%22%3A%22%22%2C%22DOI%22%3A%2210.1017%5C%2Fjog.2020.83%22%2C%22ISSN%22%3A%220022-1430%22%2C%22url%22%3A%22%22%2C%22collections%22%3A%5B%22D2UAFEX5%22%2C%22EY7L9SEU%22%5D%2C%22dateModified%22%3A%222022-08-15T17%3A42%3A00Z%22%7D%7D%5D%7D
Jenkins, W. F., Gerstoft, P., & Park, Y. (2024). Geoacoustic inversion using Bayesian optimization with a Gaussian process surrogate model. The Journal of the Acoustical Society of America, 156(2), 812–822. https://doi.org/10.1121/10.0028177
Zhou, Z., Gerstoft, P., & Olsen, K. (2024). Graph-learning approach to combine multiresolution seismic velocity models. Geophysical Journal International, 238(3), 1353–1365. https://doi.org/10.1093/gji/ggae212
Li, J., Gerstoft, P., & Fan, J. (2024). Eigenvalues of the noise covariance matrix in ocean waveguides. The Journal of the Acoustical Society of America, 156(1), 189–201. https://doi.org/10.1121/10.0026477
Liu, R., & Gerstoft, P. (2024). Spatial acoustic properties recovery with deep learning. The Journal of the Acoustical Society of America, 155(6), 3690–3701. https://doi.org/10.1121/10.0026231
Zhou, Z., Gerstoft, P., & Olsen, K. B. (2024). 3D Multiresolution Velocity Model Fusion with Probability Graphical Models. Bulletin of the Seismological Society of America, 114(3), 1279–1292. https://doi.org/10.1785/0120230271
Wiens, D. A., Aster, R. C., Nyblade, A. A., Bromirski, P. D., Gerstoft, P., & Stephen, R. A. (2024). Ross Ice Shelf Displacement and Elastic Plate Waves Induced by Whillans Ice Stream Slip Events. Geophysical Research Letters, 51(7), e2023GL108040. https://doi.org/10.1029/2023GL108040
Yoon, S., Park, Y., Gerstoft, P., & Seong, W. (2024). Predicting ocean pressure field with a physics-informed neural network. The Journal of the Acoustical Society of America, 155(3), 2037–2049. https://doi.org/10.1121/10.0025235
Mecklenbräuker, C. F., Gerstoft, P., Ollila, E., & Park, Y. (2024). Robust and sparse M-estimation of DOA. Signal Processing, 220, 109461. https://doi.org/10.1016/j.sigpro.2024.109461
Park, Y., & Gerstoft, P. (2024). Difference Frequency Gridless Sparse Array Processing. IEEE Open Journal of Signal Processing, 5, 914–925. https://doi.org/10.1109/OJSP.2024.3425284
Wu, Y., Wakin, M. B., & Gerstoft, P. (2024). Non-Uniform Array and Frequency Spacing for Regularization-Free Gridless DOA. IEEE Transactions on Signal Processing, 72, 2006–2020. https://doi.org/10.1109/TSP.2024.3386018
Sathyanarayanan, V., Gerstoft, P., & El Gamal, A. (2023). RML22: Realistic Dataset Generation for Wireless Modulation Classification. Ieee Transactions on Wireless Communications, 22(11), 7663–7675. https://doi.org/10.1109/Twc.2023.3254490
Jenkins, W. F., Gerstoft, P., & Park, Y. (2023). Bayesian optimization with Gaussian process surrogate model for source localization. The Journal of the Acoustical Society of America, 154(3), 1459–1470. https://doi.org/10.1121/10.0020839
Khurjekar, I. D., & Gerstoft, P. (2023). Uncertainty quantification for direction-of-arrival estimation with conformal prediction. The Journal of the Acoustical Society of America, 154(2), 979–990. https://doi.org/10.1121/10.0020655
Lee, J.-H., Park, Y., & Gerstoft, P. (2023). Compressive frequency-difference direction-of-arrival estimation. The Journal of the Acoustical Society of America, 154(1), 141–151. https://doi.org/10.1121/10.0020053
Liu, Y., Zhao, Y., Gerstoft, P., Zhou, F., Qiao, G., & Yin, J. (2023). Deep transfer learning-based variable Doppler underwater acoustic communications. The Journal of the Acoustical Society of America, 154(1), 232–244. https://doi.org/10.1121/10.0020147
Liu, R., Gerstoft, P., Bianco, M. J., & Rao, B. D. (2023). Recovery of spatially varying acoustical properties via automated partial differential equation identification. The Journal of the Acoustical Society of America, 153(6), 3169. https://doi.org/10.1121/10.0019592
Michalopoulou, Z. H., & Gerstoft, P. (2023). Waveform modeling with Gaussian Processes for inversion in ocean acoustics. Journal of the Acoustical Society of America, 153(3). https://doi.org/10.1121/10.0018577
Michalopoulou, Z. H., & Gerstoft, P. (2023). Inversion in an uncertain ocean using Gaussian processes. Journal of the Acoustical Society of America, 153(3), 1600–1611. https://doi.org/10.1121/10.0017437
Jenkins, W. F., Gerstoft, P., Chien, C. C., & Ozanich, E. (2023). Reducing dimensionality of spectrograms using convolutional autoencoders. Journal of the Acoustical Society of America, 153(3). https://doi.org/10.1121/10.0018582
Chien, C.-C., Jenkins, W. F., Gerstoft, P., Zumberge, M., & Mellors, R. (2023). Automatic classification with an autoencoder of seismic signals on a distributed acoustic sensing cable. Computers and Geotechnics, 155, 105223. https://doi.org/10.1016/j.compgeo.2022.105223
Fernandez-Grande, E., Karakonstantis, X., Caviedes-Nozal, D., & Gerstoft, P. (2023). Generative models for sound field reconstruction. The Journal of the Acoustical Society of America, 153(2), 1179–1190. https://doi.org/10.1121/10.0016896
Park, Y., Meyer, F., & Gerstoft, P. (2023). Graph-based sequential beamforming. The Journal of the Acoustical Society of America, 153(1), 723–737. https://doi.org/10.1121/10.0016876
De Salvio, D., Bianco, M. J., Gerstoft, P., D’Orazio, D., & Garai, M. (2023). Blind source separation by long-term monitoring: A variational autoencoder to validate the clustering analysis. The Journal of the Acoustical Society of America, 153(1), 738–750. https://doi.org/10.1121/10.0016887
Wu, Y. F., Wakin, M. B., & Gerstoft, P. (2023). Gridless DOA Estimation With Multiple Frequencies. Ieee Transactions on Signal Processing, 71, 417–432. https://doi.org/10.1109/Tsp.2023.3244091
Chaput, J., Aster, R., Karplus, M., Nakata, N., Gerstoft, P., Bromirski, P. D., Nyblade, A., Stephen, R. A., & Wiens, D. A. (2022). Near-surface seismic anisotropy in Antarctic glacial snow and ice revealed by high-frequency ambient noise. Journal of Glaciology, 1–17. https://doi.org/10.1017/jog.2022.98
Collins, M. D., Turgut, A., Buckingham, M. J., Gerstoft, P., & Siderius, M. (2022). Selected Topics of the Past Thirty Years in Ocean Acoustics. Journal of Theoretical and Computational Acoustics, 30(03), 22. https://doi.org/10.1142/s2591728522400011
Michalopoulou, Z. H., Gerstoft, P., Rios, D., & Hodgkiss, W. S. (2022). Tracking and Inversion Using Midfrequency Signals in the Seabed Characterization Experiment. Ieee Journal of Oceanic Engineering, 47(3), 657–669. https://doi.org/10.1109/Joe.2021.3122284
Park, Y., & Gerstoft, P. (2022). Gridless sparse covariance-based beamforming via alternating projections including co-prime arrays. Journal of the Acoustical Society of America, 151(6), 3828–3837. https://doi.org/10.1121/10.0011617
Gerstoft, P., Hu, Y. H., Bianco, M. J., Patil, C., Alegre, A., Freund, Y., & Grondin, F. (2022). Audio scene monitoring using redundant ad hoc microphone array networks. Ieee Internet of Things Journal, 9(6), 4259–4268. https://doi.org/10.1109/jiot.2021.3103523
Zhou, Z., Bianco, M., Gerstoft, P., & Olsen, K. (2022). High-Resolution Imaging of Complex Shallow Fault Zones Along the July 2019 Ridgecrest Ruptures. Geophysical Research Letters, 49(1). https://doi.org/ARTN e2021GL095024 10.1029/2021GL095024
Whiteaker, B., & Gerstoft, P. (2022). Reducing echo state network size with controllability matrices. Chaos: An Interdisciplinary Journal of Nonlinear Science, 32(7), 073116. https://doi.org/10.1063/5.0071926
Hahmann, M., Fernandez-Grande, E., Gunawan, H., & Gerstoft, P. (2022). Sound source localization using multiple ad hoc distributed microphone arrays. JASA Express Letters, 2(7), 074801. https://doi.org/10.1121/10.0011811
Herchig, R., Palermo, N., Gerstoft, P., Daniel, T., & Sternlicht, D. (2022). Comparing the Performance of Convolutional Neural Networks Trained to Localize Underwater Sound Sources. 2022 Oceans Hampton Roads. https://doi.org/10.1109/Oceans47191.2022.9977120
Bell, R., Harris, F. J., Gerstoft, P., & Bharadia, D. (2022). High Resolution Spectral Analysis and Signal Segregation Using the Polyphase Channelizer. 2022 56th Asilomar Conference on Signals, Systems, and Computers, 519–526. https://doi.org/10.1109/Ieeeconf56349.2022.10051908
Park, Y., Gerstoft, P., & Lee, J.-H. (2022). Difference-Frequency MUSIC for DOAs. IEEE Signal Processing Letters, 29, 2612–2616. https://doi.org/10.1109/LSP.2022.3230365
Groll, H., Gerstoft, P., Hofer, M., Blumenstein, J., Zemen, T., & Mecklenbrauker, C. F. (2022). Scatterer Identification by Atomic Norm Minimization in Vehicular mm-Wave Propagation Channels. Ieee Access, 10, 102334–102354. https://doi.org/10.1109/access.2022.3205616
Lai, P., Amirkulova, F., & Gerstoft, P. (2021). Conditional Wasserstein generative adversarial networks applied to acoustic metamaterial design. Journal of the Acoustical Society of America, 150(6), 4362–4374. https://doi.org/10.1121/10.0008929
Liu, R. X., Bianco, M. J., & Gerstoft, P. (2021). Automated partial differential equation identification. Journal of the Acoustical Society of America, 150(4), 2364–2374. https://doi.org/10.1121/10.0006444
Aster, R. C., Lipovsky, B. P., Cole, H. M., Bromirski, P. D., Gerstoft, P., Nyblade, A., Wiens, D. A., & Stephen, R. (2021). Swell-triggered seismicity at the near-front damage zone of the Ross Ice Shelf. Seismological Research Letters, 92(5), 2768–2792. https://doi.org/10.1785/0220200478
Jenkins, W. F., Gerstoft, P., Bianco, M. J., & Bromirski, P. D. (2021). Unsupervised deep clustering of seismic data: Monitoring the Ross Ice Shelf, Antarctica. Journal of Geophysical Research-Solid Earth, 126(9), 26. https://doi.org/10.1029/2021jb021716
Li, J., Siderius, M., Gerstoft, P., Fan, J., & Muzi, L. (2021). Head-wave correlations in layered seabed: Theory and modeling. JASA Express Letters, 1(9), 7. https://doi.org/10.1121/10.0006390
Shah, T., Zhuo, L., Lai, P., De la Rosa-Moreno, A., Amirkulova, F., & Gerstoft, P. (2021). Reinforcement learning applied to metamaterial design. Journal of the Acoustical Society of America, 150(1), 321–338. https://doi.org/10.1121/10.0005545
Niu, H. Q., Gerstoft, P., Zhang, R. H., Li, Z. L., Gong, Z. X., & Wang, H. B. (2021). Mode separation with one hydrophone in shallow water: A sparse Bayesian learning approach based on phase speed. Journal of the Acoustical Society of America, 149(6), 4366–4376. https://doi.org/10.1121/10.0005312
Wagner, M., Park, Y., & Gerstoft, P. (2021). Gridless DOA estimation and root-MUSIC for non-uniform linear arrays. Ieee Transactions on Signal Processing, 69, 2144–2157. https://doi.org/10.1109/tsp.2021.3068353
Ozanich, E., Thode, A., Gerstoft, P., Freeman, L. A., & Freeman, S. (2021). Deep embedded clustering of coral reef bioacoustics. Journal of the Acoustical Society of America, 149(4), 2587–2601. https://doi.org/10.1121/10.0004221
Park, Y., Meyer, F., & Gerstoft, P. (2021). Sequential sparse Bayesian learning for time-varying direction of arrival. Journal of the Acoustical Society of America, 149(3), 2089–2099. https://doi.org/10.1121/10.0003802
Cao, H. G., Wang, W. B., Su, L., Ni, H. Y., Gerstoft, P., Ren, Q. Y., & Ma, L. (2021). Deep transfer learning for underwater direction of arrival using one vector sensora). Journal of the Acoustical Society of America, 149(3), 1699–1711. https://doi.org/10.1121/10.0003645
Snover, D., Johnson, C. W., Bianco, M. J., & Gerstoft, P. (2021). Deep clustering to identify sources of urban seismic noise in Long Beach, California. Seismological Research Letters, 92(2), 1011–1022. https://doi.org/10.1785/0220200164
Thode, A. M., Conrad, A. S., Ozanich, E., King, R., Freeman, S. E., Freeman, L. A., Zgliczynski, B., Gerstoft, P., & Kim, K. H. (2021). Automated two-dimensional localization of underwater acoustic transient impulses using vector sensor image processing (vector sensor localization). Journal of the Acoustical Society of America, 149(2), 770–787. https://doi.org/10.1121/10.0003382
Baker, M. G., Aster, R. C., Wiens, D. A., Nyblade, A., Bromirski, P. D., Gerstoft, P., & Stephen, R. A. (2021). Teleseismic earthquake wavefields observed on the Ross Ice Shelf. Journal of Glaciology, 67(261), 58–74. https://doi.org/10.1017/jog.2020.83