PseudoProp: Robust Pseudo-Label Generation for Semi-Supervised Object Detection in Autonomous Driving Systems

Shu Hu, Chun-Hao Liu, Jayanta Dutta, Ming-Ching Chang, Siwei Lyu, Naveen Ramakrishnan

CVPRW 2022 pp. 4389-4397

doi:10.1109/CVPRW56347.2022.00485 /cvprw/2022/hu2022cvprw-pseudoprop/

Abstract

Semi-supervised object detection methods are widely used in autonomous driving systems, where only a fraction of objects are labeled. To propagate information from the labeled objects to the unlabeled ones, pseudo-labels for unlabeled objects must be generated. Although pseudo-labels have proven to improve the performance of semi-supervised object detection significantly, the applications of image-based methods to video frames result in numerous miss or false detections using such generated pseudo-labels. In this paper, we propose a new approach, Pseudo-Prop, to generate robust pseudo-labels by leveraging motion continuity in video frames. Specifically, PseudoProp uses a novel bidirectional pseudo-label propagation approach to compensate for misdetection. A feature-based fusion technique is also used to suppress inference noise. Extensive experiments on the large-scale Cityscapes dataset demonstrate that our method outperforms the state-of-the-art semi-supervised object detection methods by 7.4% on mAP75.

PDF CVPRW Semantic Scholar

Cite

Text

Hu et al. "PseudoProp: Robust Pseudo-Label Generation for Semi-Supervised Object Detection in Autonomous Driving Systems." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2022. doi:10.1109/CVPRW56347.2022.00485

Markdown

[Hu et al. "PseudoProp: Robust Pseudo-Label Generation for Semi-Supervised Object Detection in Autonomous Driving Systems." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2022.](https://mlanthology.org/cvprw/2022/hu2022cvprw-pseudoprop/) doi:10.1109/CVPRW56347.2022.00485

BibTeX

@inproceedings{hu2022cvprw-pseudoprop,
  title     = {{PseudoProp: Robust Pseudo-Label Generation for Semi-Supervised Object Detection in Autonomous Driving Systems}},
  author    = {Hu, Shu and Liu, Chun-Hao and Dutta, Jayanta and Chang, Ming-Ching and Lyu, Siwei and Ramakrishnan, Naveen},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
  year      = {2022},
  pages     = {4389-4397},
  doi       = {10.1109/CVPRW56347.2022.00485},
  url       = {https://mlanthology.org/cvprw/2022/hu2022cvprw-pseudoprop/}
}