Passive acoustic monitoring has been rapidly scaled up to augment traditional visual surveys. While signal processing tools can be used to analyze large batches of acoustic data they cannot be used to find specific sounds, such as diverse and abundant fish sounds. A large bottleneck has been the human effort to find and label these call types. I collaborated with the Mooney lab by training a YOLOv5 convolutional neural network to automate the detection of tonal and pulsed fish calls in spectrogram data from five tropical coral reefs in the U.S. Virgin Islands. This also helped train a neural network on a floating mooring broadcasting real-time updates on fish call counts. Collaborator