English

SecondPose: SE(3)-Consistent Dual-Stream Feature Fusion for Category-Level Pose Estimation

Computer Vision and Pattern Recognition 2024-03-25 v3

Abstract

Category-level object pose estimation, aiming to predict the 6D pose and 3D size of objects from known categories, typically struggles with large intra-class shape variation. Existing works utilizing mean shapes often fall short of capturing this variation. To address this issue, we present SecondPose, a novel approach integrating object-specific geometric features with semantic category priors from DINOv2. Leveraging the advantage of DINOv2 in providing SE(3)-consistent semantic features, we hierarchically extract two types of SE(3)-invariant geometric features to further encapsulate local-to-global object-specific information. These geometric features are then point-aligned with DINOv2 features to establish a consistent object representation under SE(3) transformations, facilitating the mapping from camera space to the pre-defined canonical space, thus further enhancing pose estimation. Extensive experiments on NOCS-REAL275 demonstrate that SecondPose achieves a 12.4% leap forward over the state-of-the-art. Moreover, on a more complex dataset HouseCat6D which provides photometrically challenging objects, SecondPose still surpasses other competitors by a large margin.

Keywords

Cite

@article{arxiv.2311.11125,
  title  = {SecondPose: SE(3)-Consistent Dual-Stream Feature Fusion for Category-Level Pose Estimation},
  author = {Yamei Chen and Yan Di and Guangyao Zhai and Fabian Manhardt and Chenyangguang Zhang and Ruida Zhang and Federico Tombari and Nassir Navab and Benjamin Busam},
  journal= {arXiv preprint arXiv:2311.11125},
  year   = {2024}
}

Comments

CVPR 2024 accepted. Code is available at: https://github.com/NOrangeeroli/SecondPose

R2 v1 2026-06-28T13:25:07.482Z