Roughly 40 million images of the sea floor have been collected by the NOAA Habcam vehicle since 2012. Only about 2% of these have been annotated by humans for objects of interest such as fish, scallops, and other invertebrates. NOAA has been developing a system (VIAME, see www.viametoolkit.org) to automatically detect objects of interest in underwater images using artificial intelligence. In this project, the intern would help train the VIAME system to automatically detect targets of interest (e.g., fish species, sea stars, sponges), and then evaluate the performance of the trained system. There may be an opportunity for the intern to go to sea to help collect data for this work. See this article for more information about this type of work.