MonoPair: Monocular 3D Object Detection Using Pairwise Spatial Relationships

TL;DR

MonoPair enhances monocular 3D object detection by leveraging pairwise spatial relationships and uncertainty modeling, significantly improving detection accuracy for occluded objects in autonomous driving scenarios.

Contribution

It introduces a novel pairwise relationship modeling approach with uncertainty-aware predictions and joint optimization, advancing monocular 3D detection performance.

Findings

Outperforms state-of-the-art on KITTI benchmark

Significantly improves detection of occluded objects

Maintains run-time efficiency with integrated structure

Abstract

Monocular 3D object detection is an essential component in autonomous driving while challenging to solve, especially for those occluded samples which are only partially visible. Most detectors consider each 3D object as an independent training target, inevitably resulting in a lack of useful information for occluded samples. To this end, we propose a novel method to improve the monocular 3D object detection by considering the relationship of paired samples. This allows us to encode spatial constraints for partially-occluded objects from their adjacent neighbors. Specifically, the proposed detector computes uncertainty-aware predictions for object locations and 3D distances for the adjacent object pairs, which are subsequently jointly optimized by nonlinear least squares. Finally, the one-stage uncertainty-aware prediction structure and the post-optimization module are dedicatedly integrated for ensuring the run-time efficiency. Experiments demonstrate that our method yields the best performance on KITTI 3D detection benchmark, by outperforming state-of-the-art competitors by wide margins, especially for the hard samples.