Reproduce “Exploiting Sample Uncertainty for Domain Adaptive Person Re-Identification” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Kecheng Zheng, Cuiling Lan, Wenjun Zeng, Zhizheng Zhang, Zheng-Jun Zha
UNRN adapts person re-identification across datasets by reducing the influence of uncertain target pseudo-labels. It compares soft-label distributions from a student and mean-teacher against source-class and target-cluster centers, then uses disagreement to weight classification, triplet, and contrastive losses. On Duke-to-Market, it achieves 78.1% mAP versus MEB-Net’s 76.0%, with highest mAP in four transfers; this does not establish perfect error detection.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.