TransPose: 6D Object Pose Estimation with Geometry-Aware Transformer

arXiv:2310.16279v2 Announce Type: replace
Abstract: Estimating the 6D object pose is an essential task in many applications. Due to the lack of depth information, existing RGB-based methods are sensitive to occlusion and illumination changes. How to extract and utilize the geometry features in depth information is crucial to achieve accurate predictions. To this end, we propose TransPose, a novel 6D pose framework that exploits Transformer Encoder with geometry-aware module to develop better learning of point cloud feature representations. Specifically, we …

abstract applications arxiv cs.cv extract features geometry information object predictions transformer type