papersTODAY 04:00 UTC
Deep learning segmentation framework detects fin whale calls in ocean-bottom seismic data
Researchers present a semantic segmentation deep learning framework for large-scale bioacoustic detection, applied to fin whale calls recorded by ocean-bottom seismometers. The approach repurposes seismic instruments, originally deployed for geophysics, as a broad-area passive acoustic monitoring network. This could expand monitoring of baleen whales over long time spans and wide ocean regions.