This project combines multiple foundational models for detection, segmentation, mesh-based inference, 3D posture detection, and temporal modeling in canonical space to unify rodent behavioral analysis. Our ultimate goal is to support multi-domain integration for inferential discovery.
Overview
Our foundational models are supported by a vast and highly curated rodent image dataset. Interact with some representative images below, and test out performance in model demos.
Posture
Hover to view.
Dataset summary.
Use this section for a short description of the first project component.
Use this section for a short description of the first project component.
Behavior
Hover to view.
Behavior summary.
Use this section for a short description of the first project component.
Use this section for a short description of the first project component.