Transparent Object Depth Completion
This addresses a major challenge in robotic grasp and manipulation for transparent objects, which is an incremental improvement over existing depth completion methods.
The paper tackles the problem of inaccurate depth maps for transparent objects in robotic perception by proposing an end-to-end network that combines single-view and multi-view depth estimation with a refinement module, achieving superior accuracy and robustness on ClearPose and TransCG datasets compared to state-of-the-art methods.
The perception of transparent objects for grasp and manipulation remains a major challenge, because existing robotic grasp methods which heavily rely on depth maps are not suitable for transparent objects due to their unique visual properties. These properties lead to gaps and inaccuracies in the depth maps of the transparent objects captured by depth sensors. To address this issue, we propose an end-to-end network for transparent object depth completion that combines the strengths of single-view RGB-D based depth completion and multi-view depth estimation. Moreover, we introduce a depth refinement module based on confidence estimation to fuse predicted depth maps from single-view and multi-view modules, which further refines the restored depth map. The extensive experiments on the ClearPose and TransCG datasets demonstrate that our method achieves superior accuracy and robustness in complex scenarios with significant occlusion compared to the state-of-the-art methods.