Text Based Person Retrieval With Noisy 1
評価指標
Rank 1
Rank-10
Rank-5
mAP
mINP
評価結果
このベンチマークにおける各モデルのパフォーマンス結果
モデル名 | Rank 1 | Rank-10 | Rank-5 | mAP | mINP | Paper Title | Repository |
|---|---|---|---|---|---|---|---|
| SSAN | 40.57 | 71.53 | 62.58 | 20.93 | 2.22 | Semantically Self-Aligned Network for Text-to-Image Part-aware Person Re-identification | |
| RDE | 66.54 | 86.70 | 81.70 | 39.08 | 7.55 | Noisy-Correspondence Learning for Text-to-Image Person Re-identification | |
| CLIP-C | 55.25 | 81.32 | 74.76 | 31.09 | 4.94 | Learning Transferable Visual Models From Natural Language Supervision | |
| IVT | 50.21 | 76.18 | 69.14 | 34.72 | 8.77 | See Finer, See More: Implicit Modality Alignment for Text-based Person Retrieval | |
| IRRA | 60.76 | 84.01 | 78.26 | 35.87 | 6.80 | Cross-Modal Implicit Relation Reasoning and Aligning for Text-to-Image Person Retrieval | - |
| DECL | 61.95 | 83.88 | 78.36 | 36.08 | 6.25 | Deep Evidential Learning with Noisy Correspondence for Cross-Modal Retrieval |
0 of 6 row(s) selected.