Research map / Label-efficient and robust learning / Recognition and long-tail
Syncretic Modality Collaborative Learning for Visible Infrared Person Re-Identification
Ziyu Wei, Xi Yang, Nannan Wang, Xinbo Gao
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 challenge enhanced homogeneity learning (CEHL) (ablation) | baseline with SMGM · SYSU-MM01 / RegDB | mAP: +0.52% on SYSU-MM01, +2.88% on RegDB | Improved |
| add auxiliary distributional similarity learning (ADSL) (ablation) | baseline with CEHL · SYSU-MM01 | rank-1 and mAP: +3.19% rank-1, +3.91% mAP | Improved |
| add incremental training (IT) strategy (ablation) | SMCL without IT · SYSU-MM01 | rank-1: 63.16 -> 67.39 | Improved |
| add incremental training (IT) strategy (ablation) | SMCL without IT · RegDB | mAP: 73.85 -> 78.57 | Improved |
| add an auxiliary modality of any kind (ablation) | SMCL · SYSU-MM01 | mAP: at least 4.64% higher | Improved |
| use syncretic modality instead of grayscale modality (ablation) | SMCL · SYSU-MM01 | rank-1: +3.16% | Improved |
| use syncretic modality instead of X modality (ablation) | SMCL · RegDB | mAP: +5.57% | Improved |
| use grayscale modality instead of no auxiliary modality (ablation) | SMCL · RegDB | rank-1: 78.49 -> 74.36 | Hurt |
| use grayscale modality instead of X modality (ablation) | SMCL · SYSU-MM01 vs RegDB | rank-1: 64.23 vs 62.34 on SYSU-MM01; 74.36 vs 79.46 on RegDB | Mixed |
| set CEHL difficulty margin m to 4 instead of 0 (ablation) | CEHL · SYSU-MM01 | rank-1: 67.39 | Improved |
| use large margin m instead of m=1 (ablation) | CEHL · RegDB | rank-1: 83.05 at m=1 | Hurt |
| use ADSL instead of batch-hard triplet loss (ablation) | metric learning · SYSU-MM01 | rank-1: 61.53 -> 67.39 | Improved |
Closest papers by meaning
- Modality Unifying Network for Visible-Infrared Person Re-Identification (ICCV 2023 · similarity 0.94)
- Modality-Aware Bias Mitigation and Invariance Learning for Unsupervised Visible-Infrared Person Re-Identification (AAAI 2026 · similarity 0.93)
- Joint Color-irrelevant Consistency Learning and Identity-aware Modality Adaptation for Visible-infrared Cross Modality Person Re-identification (AAAI 2021 · similarity 0.93)
- Implicit Discriminative Knowledge Learning for Visible-Infrared Person Re-Identification (CVPR 2024 · similarity 0.92)
- Shallow-Deep Collaborative Learning for Unsupervised Visible-Infrared Person Re-Identification (CVPR 2024 · similarity 0.92)
- Empowering Visible-Infrared Person Re-Identification with Large Foundation Models (NeurIPS 2024 · similarity 0.92)
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.
