3DRegNet: A Deep Neural Network for 3D Point Registration
This addresses the problem of aligning 3D scans for applications like robotics or computer vision, but it is incremental as it builds on existing correspondence-based methods with a refinement network.
The paper tackles 3D point registration by introducing 3DRegNet, a deep learning architecture that classifies point correspondences as inliers/outliers and regresses motion parameters, achieving state-of-the-art results on two challenging datasets with higher speedup compared to baselines.
We present 3DRegNet, a novel deep learning architecture for the registration of 3D scans. Given a set of 3D point correspondences, we build a deep neural network to address the following two challenges: (i) classification of the point correspondences into inliers/outliers, and (ii) regression of the motion parameters that align the scans into a common reference frame. With regard to regression, we present two alternative approaches: (i) a Deep Neural Network (DNN) registration and (ii) a Procrustes approach using SVD to estimate the transformation. Our correspondence-based approach achieves a higher speedup compared to competing baselines. We further propose the use of a refinement network, which consists of a smaller 3DRegNet as a refinement to improve the accuracy of the registration. Extensive experiments on two challenging datasets demonstrate that we outperform other methods and achieve state-of-the-art results. The code is available.