Reef Fish Call Detection


Improving Reef Fish Call Detection Across Recording Devices


Passive acoustic monitoring promotes a deeper understanding of the biodiversity and health of marine ecosystems that are hard to appreciate from visual surveys alone. A large bottleneck has been the human effort required to find and label diverse and abundant fish call types in acoustic recordings. With the Mooney and WARP lab, I collaborated on improving recording devices and training machine learning models to reveal new fish call data from five tropical coral reefs in the U.S. Virgin Islands and Hawaii. These machine learning models have potential to be rapidly scaled up for continuous ecosystem monitoring, improve estimates of fish and invertebrate density and distribution, and paired with eDNA, visual, and hydrographic models to help fuel new strategies for reef restoration. Collaborator
Share
Tools
Translate to