This Project Grant award from the National Science Foundation (NSF) under the Computer and Information Science and Engineering (CISE) program (CFDA 47.070) provides $599,939 to Georgia Tech Research Corporation to develop advanced computational methods for analyzing complex multi-animal behaviors. The key products and services to be delivered include:
- New self-supervised learning techniques for segmenting long behavior sequences into distinct short-term actions like grooming or sniffing.
- A novel inverse reinforcement learning framework to explore how animals make decisions and how these decisions change over time.
- A new spatiotemporal graph model to capture interactions between animals and study short-term actions and long-term decision-making in social contexts.
These interconnected aims seek to create a practical set of accurate, interpretable, and reliable machine learning models for analyzing multi-animal behaviors in naturalistic settings. The insights gained have the potential to reveal neuro-behavioral correlations and impact fields such as AI, robotics, cognitive science, and psychology. The award period is from Oct 1, 2024 to Sep 30, 2027.
Generated 1/28/25, 7:23 AM