COTReg:Coupled Optimal Transport based Point Cloud Registration
This work addresses the critical step of correspondence prediction in point cloud registration for applications in robotics and computer vision, representing an incremental improvement over existing methods.
The paper tackles the problem of generating high-quality correspondences for 3D point cloud registration by proposing COTReg, a learning framework that jointly optimizes pointwise and structural matchings as a coupled optimal transport problem, achieving state-of-the-art performance on benchmarks like 3DMatch, KITTI, 3DCSR, and ModelNet40.
Generating a set of high-quality correspondences or matches is one of the most critical steps in point cloud registration. This paper proposes a learning framework COTReg by jointly considering the pointwise and structural matchings to predict correspondences of 3D point cloud registration. Specifically, we transform the two matchings into a Wasserstein distance-based and a Gromov-Wasserstein distance-based optimizations, respectively. Thus the task of establishing the correspondences can be naturally reshaped to a coupled optimal transport problem. Furthermore, we design a network to predict the confidence score of being an inlier for each point of the point clouds, which provides the overlap region information to generate correspondences. Our correspondence prediction pipeline can be easily integrated into either learning-based features like FCGF or traditional descriptors like FPFH. We conducted comprehensive experiments on 3DMatch, KITTI, 3DCSR, and ModelNet40 benchmarks, showing the state-of-art performance of the proposed method.