English

G2L-Net: Global to Local Network for Real-time 6D Pose Estimation with Embedding Vector Features

Computer Vision and Pattern Recognition 2020-03-27 v2

Abstract

In this paper, we propose a novel real-time 6D object pose estimation framework, named G2L-Net. Our network operates on point clouds from RGB-D detection in a divide-and-conquer fashion. Specifically, our network consists of three steps. First, we extract the coarse object point cloud from the RGB-D image by 2D detection. Second, we feed the coarse object point cloud to a translation localization network to perform 3D segmentation and object translation prediction. Third, via the predicted segmentation and translation, we transfer the fine object point cloud into a local canonical coordinate, in which we train a rotation localization network to estimate initial object rotation. In the third step, we define point-wise embedding vector features to capture viewpoint-aware information. To calculate more accurate rotation, we adopt a rotation residual estimator to estimate the residual between initial rotation and ground truth, which can boost initial pose estimation performance. Our proposed G2L-Net is real-time despite the fact multiple steps are stacked via the proposed coarse-to-fine framework. Extensive experiments on two benchmark datasets show that G2L-Net achieves state-of-the-art performance in terms of both accuracy and speed.

Keywords

Cite

@article{arxiv.2003.11089,
  title  = {G2L-Net: Global to Local Network for Real-time 6D Pose Estimation with Embedding Vector Features},
  author = {Wei Chen and Xi Jia and Hyung Jin Chang and Jinming Duan and Ales Leonardis},
  journal= {arXiv preprint arXiv:2003.11089},
  year   = {2020}
}

Comments

10 pages, 11 figures, accepted in CVPR 2020

R2 v1 2026-06-23T14:26:02.636Z