English

CATRE: Iterative Point Clouds Alignment for Category-level Object Pose Refinement

Computer Vision and Pattern Recognition 2022-07-19 v1

Abstract

While category-level 9DoF object pose estimation has emerged recently, previous correspondence-based or direct regression methods are both limited in accuracy due to the huge intra-category variances in object shape and color, etc. Orthogonal to them, this work presents a category-level object pose and size refiner CATRE, which is able to iteratively enhance pose estimate from point clouds to produce accurate results. Given an initial pose estimate, CATRE predicts a relative transformation between the initial pose and ground truth by means of aligning the partially observed point cloud and an abstract shape prior. In specific, we propose a novel disentangled architecture being aware of the inherent distinctions between rotation and translation/size estimation. Extensive experiments show that our approach remarkably outperforms state-of-the-art methods on REAL275, CAMERA25, and LM benchmarks up to a speed of ~85.32Hz, and achieves competitive results on category-level tracking. We further demonstrate that CATRE can perform pose refinement on unseen category. Code and trained models are available.

Keywords

Cite

@article{arxiv.2207.08082,
  title  = {CATRE: Iterative Point Clouds Alignment for Category-level Object Pose Refinement},
  author = {Xingyu Liu and Gu Wang and Yi Li and Xiangyang Ji},
  journal= {arXiv preprint arXiv:2207.08082},
  year   = {2022}
}

Comments

accepted by ECCV'22

R2 v1 2026-06-25T00:58:48.001Z