Toward Efficient Generalization in 3D Human Pose Estimation via a Canonical Domain Approach

Open in new window