Summary of Pseudo Label Refinery For Unsupervised Domain Adaptation on Cross-dataset 3d Object Detection, by Zhanwei Zhang et al.
Pseudo Label Refinery for Unsupervised Domain Adaptation on Cross-dataset 3D Object Detection
by Zhanwei Zhang, Minghao Chen, Shuai Xiao, Liang Peng, Hengjia Li, Binbin Lin, Ping Li, Wenxiao Wang, Boxi Wu, Deng Cai
First submitted to arxiv on: 30 Apr 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed pseudo label refinery framework for unsupervised domain adaptation in 3D object detection (3D UDA) improves the reliability of pseudo boxes by introducing a complementary augmentation strategy. This strategy either removes unreliable box points or replaces them with high-confidence ones. Additionally, to address the quality degradation issue caused by varying point numbers across different domains, the authors generate additional proposals and align RoI features. Experimental results demonstrate that this method surpasses state-of-the-art methods on six autonomous driving benchmarks. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper proposes a new way to make computer vision models better at recognizing objects in 3D scenes from one domain (like daytime) but trained for another (like nighttime). They do this by cleaning up the “fake labels” used during training. This is important because these fake labels can be incorrect, which makes the model learn bad things. The new method removes or replaces these unreliable labels and also helps with a problem where some datasets have more points than others. This makes it work better on real-world problems like self-driving cars. |
Keywords
» Artificial intelligence » Domain adaptation » Object detection » Unsupervised
