Mahalanobis k-NN: A Statistical Lens for Robust Point-Cloud Registrations
This work addresses robust point cloud registration for computer vision applications, presenting an incremental improvement by integrating a statistical method into existing frameworks.
The paper tackles the challenge of feature matching in learning-based point cloud registration under varying point cloud densities by using Mahalanobis k-NN to capture local neighborhood distributions, resulting in a 20% improvement in average accuracy for point cloud few-shot classification tasks.
In this paper, we discuss Mahalanobis k-NN: A Statistical Lens designed to address the challenges of feature matching in learning-based point cloud registration when confronted with an arbitrary density of point clouds. We tackle this by adopting Mahalanobis k-NN's inherent property to capture the distribution of the local neighborhood and surficial geometry. Our method can be seamlessly integrated into any local-graph-based point cloud analysis method. In this paper, we focus on two distinct methodologies: Deep Closest Point (DCP) and Deep Universal Manifold Embedding (DeepUME). Our extensive benchmarking on the ModelNet40 and FAUST datasets highlights the efficacy of the proposed method in point cloud registration tasks. Moreover, we establish for the first time that the features acquired through point cloud registration inherently can possess discriminative capabilities. This is evident by a substantial improvement of about 20% in the average accuracy observed in the point cloud few-shot classification task, benchmarked on ModelNet40 and ScanObjectNN.