Learning a Category-level Object Pose Estimator without Pose Annotations

DiffM
Re-assign community