AlphaTracker

AlphaTracker performs multi-animal tracking and pose estimation using computer vision to quantify social and individual behaviors in neuroscience experiments.

Key Features:

• Multi-Animal Tracking: Accurately tracks multiple unmarked animals simultaneously to enable analysis of social interactions.
• Pose Estimation: Estimates body-part positions to analyze the physical positioning and movements of each animal within group settings.
• Unsupervised Behavioral Clustering: Applies unsupervised learning algorithms to cluster behavioral patterns without prior labels, facilitating discovery of novel behaviors.

Scientific Applications:

• System Neuroscience: Enables quantitative analysis of social behavior and interaction dynamics in neuroscience research.
• Social Behavior Studies in Mice: Applied to studying socially-interacting mice to extract interaction patterns and behavioral motifs not captured by single-animal tracking.

Methodology:

Leverages state-of-the-art computer vision algorithms for high-accuracy multi-animal tracking, incorporates pose estimation, and employs unsupervised learning algorithms for behavioral clustering.