Deep Dive into Gradients: Better Optimization for 3D Object Detection with Gradient-Corrected IoU Supervision

Qi Ming*, Lingjuan Miao, Zhe Ma, Lin Zhao, Zhiqiang Zhou*, Xuhui Huang, Yuanpei Chen, Yufei Guo*

*Corresponding author for this work

Research output: Contribution to journalConference articlepeer-review

8 Citations (Scopus)

Abstract

Intersection-over-Union (IoU) is the most popular metric to evaluate regression performance in 3D object detection. Recently, there are also some methods applying IoU to the optimization of 3D bounding box regression. However, we demonstrate through experiments and mathematical proof that the 3D IoU loss suffers from abnormal gradient w.r.t. angular error and object scale, which further leads to slow convergence and suboptimal regression process, respectively. In this paper, we propose a Gradient-Corrected IoU (GCIoU) loss to achieve fast and accurate 3D bounding box regression. Specifically, a gradient correction strategy is designed to endow 3D IoU loss with a reasonable gradient. It ensures that the model converges quickly in the early stage of training, and helps to achieve fine-grained refinement of bounding boxes in the later stage. To solve suboptimal regression of 3D IoU loss for objects at different scales, we introduce a gradient rescaling strategy to adaptively optimize the step size. Finally, we integrate GCIoU Loss into multiple models to achieve stable performance gains and faster model convergence. Experiments on KITTI dataset demonstrate superiority of the proposed method.

Original languageEnglish
Pages (from-to)5136-5145
Number of pages10
JournalProceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition
Volume2023-June
DOIs
Publication statusPublished - 2023
Event2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2023 - Vancouver, Canada
Duration: 18 Jun 202322 Jun 2023

Keywords

  • Autonomous driving

Fingerprint

Dive into the research topics of 'Deep Dive into Gradients: Better Optimization for 3D Object Detection with Gradient-Corrected IoU Supervision'. Together they form a unique fingerprint.

Cite this