Research map / Vision-language and multimodal / Cross-modal learning
Cross-Modal Semantic Decoupling and Transfer for Text-to-Visible-Infrared Person Re-Identification
Ziang Zhang, Bin Yang, Mang Ye
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 Semantic Decoupling Learning (SDL) (ablation) | RDE baseline (no HAM init) · SYSU-TVI text-to-infrared | Rank-1 (%): 52.20 -> 55.30 | Improved |
| add SDL (ablation) | RDE with HAM init · SYSU-TVI text-to-infrared | Rank-1 (%): 59.06 -> 63.34 | Improved |
| add SDL (ablation) | RDE baseline · SYSU-TVI text-to-visible | Rank-1 (%): 90.91 -> 89.77 | No clear effect |
| add Semantic Distribution Transfer (SDT) (ablation) | SDL with HAM init · SYSU-TVI text-to-infrared | Rank-1 (%): 63.34 -> 66.21 | Improved |
| add SDT (ablation) | SDL without HAM init · SYSU-TVI text-to-infrared | Rank-1 (%): 55.30 -> 57.14 | Improved |
| add SDT (ablation) | SDL with HAM init · SYSU-TVI text-to-visible | Rank-1 (%): 95.70 -> 96.13 | Improved |
| add channel augmentation for color-irrelevant branch (ablation) | SDL · SYSU-TVI text-to-infrared | Rank-1 (%): 60.98 -> 63.34 | Improved |
| add orthogonal loss between subspaces (ablation) | SDL · SYSU-TVI text-to-infrared | mAP (%): 41.97 -> 46.25 | Improved |
| randomly replace and mask color words in text (ablation) | SDL · SYSU-TVI text-to-infrared | Rank-1 (%): 63.34 -> 63.03 | No clear effect |
| initialize with HAM TI-ReID pretrained weights (ablation) | RDE · SYSU-TVI text-to-infrared | Rank-1 (%): 52.20 -> 59.06 | Improved |
| use too large or too small alpha in SDT (ablation) | SDT · SYSU-TVI | Rank-1 | Hurt |
Closest papers by meaning
- Empowering Visible-Infrared Person Re-Identification with Large Foundation Models (NeurIPS 2024 · similarity 0.92)
- Visible-Infrared Person Re-Identification via Semantic Alignment and Affinity Inference (ICCV 2023 · similarity 0.91)
- Implicit Discriminative Knowledge Learning for Visible-Infrared Person Re-Identification (CVPR 2024 · similarity 0.90)
- Domain Shifting: A Generalized Solution for Heterogeneous Cross-Modality Person Re-Identification (ECCV 2024 · similarity 0.90)
- Unsupervised Visible-Infrared Person Re-identification under Unpaired Settings (ICCV 2025 · similarity 0.90)
- TokenMatcher: Diverse Tokens Matching for Unsupervised Visible-Infrared Person Re-Identification (AAAI 2025 · similarity 0.90)
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.
