:夫妇:联合区分性和生成性学习用于人重新认同。CVPR ' 19(口头):夫妇:
dukemtmc-reid 开源项目推荐(10 个)
dukemtmc-reid 话题在 GitHub 上没有填写简介,这里按 Star 数汇总最相关的热门仓库。
热门项目 全部 10
GitHub 共 10 个仓库,已全部展示TNNLS'22 参数高效的三维空间人员重新识别
State-of-the-art model for person re-identification in Multi-camera Multi-Target Tracking. Benchmarked on Market-1501 and DukeMTMTC-reID datasets.
A strong implementation of PCB (Beyond Part Models), outperforming all existing implementations.
Joint Disentangling and Adaptation for Cross-Domain Person Re-Identification. ECCV'20 (Oral)
DukeMTMC-reID_baseline (Matlab)
ReID + SFDA - Dual-Region Augmentation (Foreground Noise + Background Shuffle) for robust person re-identification (ReID) and source-free domain adaptation (SFDA).
A graphical user interface for labeling DukeMTMC-ReID
Person Re-ID system: ResNet-50 IBN-a + BNNeck trained on Market-1501 (Rank-1 81.80%, mAP 60.38%). FAISS gallery search, cross-dataset eval on DukeMTMC-reID, YOLOv8 video demo with persistent global IDs. Pure PyTorch, Colab T4 + Apple M1 MPS.
Repo for BMVC2020 paper "BCaR: Beginner Classifier as Regularization Towards Generalizable Re-ID"
关于这个话题
- 清单取自 GitHub 话题页并按 Star 数排序,每次展示前 10 个项目;原始页面见 https://github.com/topics/dukemtmc-reid。
- 该话题在 GitHub 上共有 10 个仓库, 本站收录的是其中最热门的 10 个,不是全量列表。
- Star / Fork 每日自动刷新一次,实时数据以 GitHub 为准。
- 当前 10 个项目累计 2044 Star,覆盖 3 种主要语言,数据更新于 2026-11-07。
- 中文介绍来自 GitHub 官方话题说明(github/explore)的机器翻译,可点「查看原文」对照;译文缺失时会先展示原文。