Research map / Vision-language and multimodal / Cross-modal learning
Dual-Teacher Interactive Knowledge Distillation Network for Text-to-Visible & Infrared Person Retrieval
Chenglong Li, Zhengyu Chen, Yifei Deng, Aihua Zheng
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 cross-view multimodal representation transfer (CVRT) (ablation) | student baseline · RGBT-PEDES | Rank-1 (%): 65.66 -> 67.51 | Improved |
| add cross-modal similarity relationship transfer (CMST) (ablation) | student baseline · RGBT-PEDES | Rank-1 (%): 65.66 -> 66.75 | Improved |
| combine CVRT and CMST (CKDM) (ablation) | student baseline · RGBT-PEDES | Rank-1 (%): 65.66 -> 69.32 | Improved |
| add IKDM on top of CKDM (ablation) | DIKDNet · RGBT-PEDES | Rank-1 (%): 69.32 -> 69.92 | Improved |
| add boundary constraint to IKDM distillation (ablation) | DIKDNet · RGBT-PEDES | Rank-1 (%): 68.79 -> 69.92 | Improved |
| IKDM without boundary constraint instead of CKDM alone (ablation) | DIKDNet · RGBT-PEDES | Rank-1 (%): 69.32 -> 68.79 | Hurt |
| use separate RGB and TIR teachers instead of a unified teacher (ablation) | DIKDNet · RGBT-PEDES | Rank-1 (%): 68.88 -> 69.92 | Improved |
| use multi-view input from different views instead of same-image views (ablation) | DIKDNet · RGBT-PEDES | Rank-1 (%): 68.43 -> 69.92 | Improved |
| replace NAM with DIKDNet | text-to-RGBT retrieval · RGBT-PEDES | Rank-1 (%): 65.51 -> 69.92 | Improved |
| replace NAM with DIKDNet | text-to-RGBT retrieval · RGBNT201-PEDES | Rank-1 (%): 43.30 -> 44.50 | Improved |
Closest papers by meaning
- Graph-Based Cross-Domain Knowledge Distillation for Cross-Dataset Text-to-Image Person Retrieval (AAAI 2025 · similarity 0.87)
- Empowering Visible-Infrared Person Re-Identification with Large Foundation Models (NeurIPS 2024 · similarity 0.85)
- Implicit Discriminative Knowledge Learning for Visible-Infrared Person Re-Identification (CVPR 2024 · similarity 0.85)
- Causality-Inspired Invariant Representation Learning for Text-Based Person Retrieval (AAAI 2024 · similarity 0.84)
- How to Make Cross Encoder a Good Teacher for Efficient Image-Text Retrieval? (CVPR 2024 · similarity 0.84)
- Let All Be Whitened: Multi-Teacher Distillation for Efficient Visual Retrieval (AAAI 2024 · similarity 0.84)
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.
