Research map / Vision-language and multimodal / CLIP and zero-shot
Person Search by Text Attribute Query As Zero-Shot Learning
Qi Dong, Shaogang Gong, Xiatian Zhu
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 local attribute-level embedding to global-only (ablation) | AIHM · Market-1501 | Rank-1 (%): 30.6 -> 43.3 | Improved |
| add local attribute-level embedding to global-only (ablation) | AIHM · DukeMTMC | Rank-1 (%): 40.7 -> 50.5 | Improved |
| add local attribute-level embedding to global-only (ablation) | AIHM · PA100K | Rank-1 (%): 26.1 -> 31.3 | Improved |
| add global category-level embedding to local-only (ablation) | AIHM · Market-1501 | Rank-1 (%): 39.5 -> 43.3 | Improved |
| add global category-level embedding to local-only (ablation) | AIHM · DukeMTMC | Rank-1 (%): 46.9 -> 50.5 | Improved |
| add global category-level embedding to local-only (ablation) | AIHM · PA100K | Rank-1 (%): 29.4 -> 31.3 | Improved |
| use quality-aware fusion instead of average pooling (ablation) | AIHM · Market-1501 | Rank-1 (%): 39.0 -> 43.3 | Improved |
| use quality-aware fusion instead of average pooling (ablation) | AIHM · DukeMTMC | Rank-1 (%): 44.9 -> 50.5 | Improved |
| use quality-aware fusion instead of average pooling (ablation) | AIHM · PA100K | Rank-1 (%): 26.1 -> 31.3 | Improved |
| add 10k negative category augmentation (ablation) | AIHM · Market-1501 | Rank-1 (%): +4.4% Rank-1 | Improved |
| increase negative augmentation from 10k to 20k (ablation) | AIHM · Market-1501, DukeMTMC, PA100K | Rank-1 (%) | Hurt |
| use CNN instead of RNN to aggregate attribute embeddings (one-hot) (ablation) | AIHM global embedding · Market-1501 | Rank-1 (%): 35.7 -> 37.1 | Improved |
Closest papers by meaning
- A Survey on Text-Based Person Search (TPAMI 2026 · similarity 0.90)
- Person Search With Natural Language Description (CVPR 2017 · similarity 0.88)
- Diverse Person: Customize Your Own Dataset for Text-Based Person Search (AAAI 2024 · similarity 0.85)
- Visual Perturbation for Text-Based Person Search (AAAI 2025 · similarity 0.85)
- An Empirical Study of CLIP for Text-Based Person Search (AAAI 2024 · similarity 0.84)
- Automatic Synthetic Data and Fine-grained Adaptive Feature Alignment for Composed Person Retrieval (NeurIPS 2025 · 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.
