High-throughput behavioral analysis for neural circuit understanding

Quantifying behavior is crucial for many applications across the life sciences and engineering. Videography provides easy methods for the observation and recording of animal behavior in diverse settings, yet extracting particular aspects of a behavior for further analysis can be highly time consuming. I will present an efficient method for markerless pose estimation based on transfer learning with deep neural networks that achieves excellent results with minimal training data. I will demonstrate the versatility of this framework by tracking various body parts in multiple species across a broad collection of behaviors from egg-laying fruits flies to hunting cheetahs. Furthermore, I will discuss new work for identifying fine-scale behaviors with deep neural networks. Lastly, I will discuss computational modeling approaches I have developed that link behavior to neural circuits.

Bio
Alexander Mathis is a Postdoctoral Fellow at Harvard University. He is interested in elucidating how the brain gives rise to adaptive behavior. For those purposes, he develops deep learning methods to analyze animal behavior, neural data, as well as creates experimentally testable computational models. His PhD thesis with Prof. Andreas Herz focused on deriving properties of grid cells from optimal coding assumptions, and figuring out how the distributed population activity can be decoded by biophysically plausible models. He was awarded a Marie Curie-Sklodowska fellowship, Human Frontiers Science Program postdoctoral fellowship, and postdoctoral fellowship by the DFG. His work was recently covered by The Atlantic & NVIDIA AI.