Learning a Category-level Object Pose Estimator without Pose Annotations

arXiv:2404.05626v1 Announce Type: new
Abstract: 3D object pose estimation is a challenging task. Previous works always require thousands of object images with annotated poses for learning the 3D pose correspondence, which is laborious and time-consuming for labeling. In this paper, we propose to learn a category-level 3D object pose estimator without pose annotations. Instead of using manually annotated images, we leverage diffusion models (e.g., Zero-1-to-3) to generate a set of images under controlled pose differences and propose to learn our …

3d object abstract annotations arxiv cs.cv estimator images labeling learn object paper type