COMPUTER VISION: A PAIR OF FRESH EYES TO STUDY ANIMAL BEHAVIOR

Computer vision has completely changed the way we study animal behavior. Over the past 15 years, researchers have developed several computer vision-based tools to automatically track an animal’s position and pose through time. This data can then be used to uncover the reasons behind the animal’s behavior. In this talk, I will compare two tracking tools, one based on contrast detection (Trilab-Tracker), and one based on machine learning (DeepLabCut), to determine which performs better on an existing set of videos of Mexican cavefish (Astyanax mexicanus) collected in Johanna Kowalko’s lab.