Object-Centric Stereo Matching for 3D Object Detection

TL;DR

This paper introduces an object-centric stereo matching approach that improves 3D object detection accuracy for autonomous driving by focusing disparity estimation on objects of interest, leading to state-of-the-art results.

Contribution

The paper proposes a novel object-centric stereo matching method that estimates disparities only for objects of interest, enhancing 3D detection accuracy over traditional disparity-focused networks.

Findings

Achieves state-of-the-art results on KITTI 3D and BEV benchmarks.

Reduces disparity errors at object boundaries, improving point cloud quality.

Outperforms existing stereo matching methods for 3D object detection.

Abstract

Safe autonomous driving requires reliable 3D object detection-determining the 6 DoF pose and dimensions of objects of interest. Using stereo cameras to solve this task is a cost-effective alternative to the widely used LiDAR sensor. The current state-of-the-art for stereo 3D object detection takes the existing PSMNet stereo matching network, with no modifications, and converts the estimated disparities into a 3D point cloud, and feeds this point cloud into a LiDAR-based 3D object detector. The issue with existing stereo matching networks is that they are designed for disparity estimation, not 3D object detection; the shape and accuracy of object point clouds are not the focus. Stereo matching networks commonly suffer from inaccurate depth estimates at object boundaries, which we define as streaking, because background and foreground points are jointly estimated. Existing networks also penalize disparity instead of the estimated position of object point clouds in their loss functions. We propose a novel 2D box association and object-centric stereo matching method that only estimates the disparities of the objects of interest to address these two issues. Our method achieves state-of-the-art results on the KITTI 3D and BEV benchmarks.