Predicting skeleton joints on egocentric video using exocentric views. This project explores cross-view transfer for human pose estimation, leveraging exocentric camera data to predict body joint positions from an ego perspective.