BEGIN:VCALENDAR VERSION:2.0 PRODID:IEEE vTools.Events//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:America/New_York BEGIN:DAYLIGHT DTSTART:20220313T030000 TZOFFSETFROM:-0500 TZOFFSETTO:-0400 RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3 TZNAME:EDT END:DAYLIGHT BEGIN:STANDARD DTSTART:20221106T010000 TZOFFSETFROM:-0400 TZOFFSETTO:-0500 RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11 TZNAME:EST END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20220612T140514Z UID:75692268-548D-4C78-A406-9A427A840030 DTSTART;TZID=America/New_York:20220609T173000 DTEND;TZID=America/New_York:20220609T183000 DESCRIPTION:The integration of passive acoustic sensors with machine learni ng enables large-scale\, low-cost\, non-invasive monitoring of vocal anima ls. A remaining challenge to the successful deployment of such monitoring systems is in maintaining high classification accuracy across complex real -world soundscapes composed of overlapping calls and variable noise patter ns. In this talk\, we present an algorithmic pipeline for using a terrestr ial co-located microphone array to (1) estimate acoustic direction-of-arri val (DoA)\, (2) distinguish individual sound sources in the environment\, and (3) approximate their spectrograms and time-domain signals. First\, we evaluate the performance of multiple DoA estimation algorithms -- includi ng an active intensity method\, white noise gain constraint beamforming\, and multiple signal classification -- as well as multiple approaches for s ource separation -- including angular thresholding\, a Gaussian mixture mo del\, and non-negative matrix factorization. Next\, we demonstrate the ana lysis pipeline for recordings collected at wildlife refuges during the daw n chorus in late spring\, when birds are most vocally active. We significa ntly improve species-level performance metrics by applying source separati on to the recordings prior to classification with the BirdNET network. Thi s approach opens possibilities for additional spatiotemporal analysis of s oundscapes\, including the ability to visualize movement and perform direc tional filtering.\n\nSpeaker(s): Irina Tolkova\, \n\nBldg: Coastal Institu te (bldg. 26)\, 215 South Ferry Rd.\, Narragansett\, Rhode Island\, United States\, 02882\, Virtual: https://events.vtools.ieee.org/m/314823 LOCATION:Bldg: Coastal Institute (bldg. 26)\, 215 South Ferry Rd.\, Narraga nsett\, Rhode Island\, United States\, 02882\, Virtual: https://events.vto ols.ieee.org/m/314823 ORGANIZER:Jason.e.gaudette@ieee.org SEQUENCE:13 SUMMARY:Spatial Bioacoustics: Soundscape Analysis with a Co-located Microph one Array URL;VALUE=URI:https://events.vtools.ieee.org/m/314823 X-ALT-DESC:Description: <br /><p>The integration of passive acoustic sensor s with machine learning enables large-scale\, low-cost\, non-invasive moni toring of vocal animals. A remaining challenge to the successful deploymen t of such monitoring systems is in maintaining high classification accurac y across complex real-world soundscapes composed of overlapping calls and variable noise patterns. In this talk\, we present an algorithmic pipeline for using a terrestrial co-located microphone array to (1) estimate acous tic direction-of-arrival (DoA)\, (2) distinguish individual sound sources in the environment\, and (3) approximate their spectrograms and time-domai n signals. First\, we evaluate the performance of multiple DoA estimation algorithms -- including an active intensity method\, white noise gain cons traint beamforming\, and multiple signal classification -- as well as mult iple approaches for source separation -- including angular thresholding\, a Gaussian mixture model\, and non-negative matrix factorization. Next\, w e demonstrate the analysis pipeline for recordings collected at wildlife r efuges during the dawn chorus in late spring\, when birds are most vocally active. We significantly improve species-level performance metrics by app lying source separation to the recordings prior to classification with the BirdNET network. This approach opens possibilities for additional spatiot emporal analysis of soundscapes\, including the ability to visualize movem ent and perform directional filtering.</p> END:VEVENT END:VCALENDAR