Research map / Label-efficient and robust learning / Self-supervised and contrastive
PASS: Part-Aware Self-Supervised Pre-training for Person Re-identification
Kuan Zhu, Haiyun Guo, Tianyi Yan, Yousong Zhu, Jinqiao Wang, Ming Tang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add [PART] tokens with part-aware matching (PASS) over DINO | ViT-S · MSMT17 (supervised) | mAP: 64.2 -> 69.1 | Improved |
| add [PART] tokens (PASS) over DINO | ViT-S · Market1501 (supervised) | mAP: 90.3 -> 92.2 | Improved |
| use PASS instead of DINO in UDA ReID | ViT-S · Market2MSMT (UDA) | mAP: 43.9 -> 49.1 | Improved |
| use PASS instead of DINO in USL ReID | ViT-S · MSMT17 (USL) | mAP: 38.4 -> 41.0 | Improved |
| use L=3 local areas instead of L=2 (ablation) | PASS ViT-S · MSMT17 (supervised) | mAP: 67.7 -> 69.1 | Improved |
| use L=3 local areas instead of L=5 (ablation) | PASS ViT-S · MSMT17 (supervised) | mAP: 67.3 -> 69.1 | Improved |
| use L=3 local areas instead of L=4 (ablation) | PASS ViT-S · Market1501 (supervised) | mAP: 91.9 -> 92.2 | Improved |
| use L=3 instead of multi-granularity L=2,3 (ablation) | PASS ViT-S · MSMT17 (supervised) | mAP: 68.7 -> 69.1 | Improved |
| concatenate [CLS] and mean [PART] instead of concatenating all tokens (ablation) | PASS ViT-S · MSMT17 (supervised) | mAP: 67.7 -> 69.1 | Improved |
| concatenate [CLS] and mean [PART] instead of the mean feature (ablation) | PASS ViT-S · MSMT17 (supervised) | mAP: 68.7 -> 69.1 | Improved |
| use mean feature instead of concatenating [CLS] and mean [PART] (ablation) | PASS ViT-S · Market2MSMT (UDA) | mAP: 44.9 -> 49.1 | Improved |
| use concatenation of [CLS] and mean [PART] instead of mean feature (ablation) | PASS ViT-S · Market1501 (USL) | Rank-1: 94.6 -> 94.9 | Mixed |
Later papers that compared with this method
5 later accepted papers report PASS as a baseline in their results tables; 5 of them report a better number on most of the results they share with it.
- Efficient Modeling of Surrogates to Improve Multi-source High-dimensional Integrative Regression (JMLR 2026 · better on 2 of 2 shared results)
- Test-time Adaptation for Foundation Medical Segmentation Model Without Parametric Updates (ICCV 2025 · better on 3 of 3 shared results)
- Instruct-ReID++: Towards Universal Purpose Instruction-Guided Person Re-Identification (TPAMI 2025 · better on 1 of 1 shared result)
- Relation-Guided Adversarial Learning for Data-Free Knowledge Transfer (IJCV 2024 · better on 1 of 1 shared result)
- MixFairFace: Towards Ultimate Fairness via MixFair Adapter in Face Recognition (AAAI 2023 · better on 2 of 3 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- A Versatile Framework for Multi-Scene Person Re-Identification (TPAMI 2024 · similarity 0.89)
- Unleashing Potential of Unsupervised Pre-Training With Intra-Identity Regularization for Person Re-Identification (CVPR 2022 · similarity 0.89)
- Instruct-ReID: A Multi-purpose Person Re-identification Task with Instructions (CVPR 2024 · similarity 0.88)
- Exploring the Camera Bias of Person Re-identification (ICLR 2025 · similarity 0.87)
- From Poses to Identity: Training-Free Person Re-Identification via Feature Centralization (CVPR 2025 · similarity 0.87)
- Multi-Prompts Learning with Cross-Modal Alignment for Attribute-Based Person Re-identification (AAAI 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
