Events

This project seeks to create a system that will analyze video footage taken from cameras attached to sharks, to better understand shark behavior in their natural habitats. Currently, researchers at OSU's Big Fish Lab spend hundreds of hours manually studying video footage to identify the ocean life that the sharks interact with, such as fish and kelp. This is a painstaking and tedious process, as a large majority of the footage is empty ocean. To help with this goal, the capstone group has trained a machine learning model to identify common ocean life in these videos without the need for human interaction. Additionally, they have created an interface to enable researchers to easily apply the model to their videos.