Facemap
Facemap is a framework for predicting neural activity from mouse orofacial movements. It includes a pose estimation model for tracking distinct keypoints on the mouse face, a neural network model for predicting neural activity using the pose estimates, and also can be used compute the singular value decomposition (SVD) of behavioral videos.
For more details, please see our paper and twitter thread.