Learning a Descriptor-Specific 3D Keypoint Detector

Keypoint detection represents the first stage in the majority of modern computer vision pipelines based on automatically established correspondences between local descriptors. However, no standard solution has emerged yet in the case of 3D data such as point clouds or meshes, which exhibit high variability in level of detail and noise. More importantly, existing proposals for 3D keypoint detection rely on geometric saliency functions that attempt to maximize repeatability rather than distinctiveness of the selected regions, which may lead to sub-optimal performance of the overall pipeline. To overcome these shortcomings, we cast 3D keypoint detection as a binary classification between points whose support can be correctly matched by a predefined 3D descriptor or not, thereby learning a descriptor-specific detector that adapts seamlessly to different scenarios. Through experiments on several public datasets, we show that this novel approach to the design of a keypoint detector represents a flexible solution that, nonetheless, can provide state-of-the-art descriptor matching performance.

1. Introduction

Detection of repeatable and distinctive keypoints is a fundamental task in modern computer vision when dealing with both images as well as 3D data. Keypoint detection in images finds applications in image retrieval, object detection and recognition, object and camera tracking, camera calibration and image registration, among the others. Key-points from 3D data are useful to deal with several shape matching tasks, such as point cloud registration, 3D object recognition and pose estimation, shape retrieval and shape classification.