Summary of Relieving Universal Label Noise For Unsupervised Visible-infrared Person Re-identification by Inferring From Neighbors, By Xiao Teng et al.
Relieving Universal Label Noise for Unsupervised Visible-Infrared Person Re-Identification by Inferring from Neighbors
by Xiao Teng, Long Lan, Dingyao Chen, Kele Xu, Nan Yin
First submitted to arxiv on: 16 Dec 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 paper proposes a novel approach for unsupervised visible-infrared person re-identification (USL-VI-ReID) by mitigating universal label noise using neighbor information. The method, called Neighbor-guided Universal Label Calibration (N-ULC), replaces explicit hard pseudo labels with soft labels derived from neighboring samples to reduce label noise. Additionally, the paper presents a Neighbor-guided Dynamic Weighting (N-DW) module to enhance training stability by minimizing the influence of unreliable samples. Experimental results on RegDB and SYSU-MM01 datasets show that the proposed method outperforms existing USL-VI-ReID approaches despite its simplicity. The approach is evaluated using benchmarks such as person re-identification metrics like CMC, rank-1, and mAP. The paper’s contribution lies in providing a straightforward yet effective solution for USL-VI-ReID, which has significant research and practical implications. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper solves a tricky problem in recognizing people from cameras that capture both visible and infrared images without any labels (annotations). Existing methods to do this often make mistakes because of noisy data. The researchers propose a new way to fix this by looking at neighboring samples to reduce errors. They also developed another technique to make the training process more stable. Tests on two large datasets show that their approach works better than other methods, even though it’s simple. This has important implications for research and real-world applications. |
Keywords
» Artificial intelligence » Unsupervised
