Fine-grained fashion image retrieval on FashionAI (test)
69.57MAP (Sleeve Length)SuperFashion
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| SuperFashionTrain Dataset=FashionAI, Evaluation Protocol=In-dataset2026.06 | 69.57 | — | 61.9 | — | — | — | 70.39 | — | 63.31 | |
| RPFMethod category=Static methods, Evaluation state=Final accuracy2026.03 | 69.15 | 66.93 | 58.83 | 72.19 | 77.14 | 72.63 | 71.48 | 71.51 | 69.38 | |
| ASENet V2+MKDMethod category=Static methods, Evaluation state=Final accuracy2026.03 | 64.22 | 69.81 | 61.31 | 73.86 | 78.51 | 74.1 | 70.67 | 68.7 | 69.41 | |
| MCL-FIRMethod category=CIL methods, Evaluation state=Final (Immediate) accuracy2026.03 | 61.28 | 64.08 | 53.25 | 68.08 | 71.68 | 70.66 | 68.43 | 62.17 | 64.41 | |
| ASENet V2+PTMethod category=Static methods, Evaluation state=Final accuracy2026.03 | 60.52 | 67.5 | 55.2 | 70.58 | 77.35 | 72.31 | 68.31 | 67.28 | 66.29 | |
| ASENet V2+GeoDCLMethod category=Static methods, Evaluation state=Final accuracy2026.03 | 59.18 | 68.71 | 55.54 | 70.72 | 77.14 | 73.03 | 68.49 | 69.25 | 66.48 | |
| ASENet V2Method category=Static methods, Evaluation state=Final accuracy2026.03 | 54.96 | 64.57 | 51.76 | 64.5 | 71.93 | 66.72 | 60.29 | 60.83 | 60.76 | |
| ASENMethod category=Static methods, Evaluation state=Final accuracy2026.03 | 53.05 | 63.04 | 52.29 | 65.26 | 71.81 | 61.39 | 63.87 | 65.25 | 61 | |
| ASENgMethod category=Static methods, Evaluation state=Final accuracy2026.03 | 50.72 | 59.95 | 47.58 | 64.32 | 68.66 | 59.39 | 55.06 | 54.73 | 56.47 | |
| CSNMethod category=Static methods, Evaluation state=Final accuracy2026.03 | 43.99 | 59.59 | 42.82 | 62.8 | 67.57 | 48.1 | 41.73 | 55.29 | 50.8 | |
| SuperFashionTrain Dataset=DARN, Evaluation Protocol=Cross-dataset2026.06 | 38.51 | — | 30.84 | — | — | — | 23.41 | — | 29.47 | |
| RPFTrain Dataset=DARN, Evaluation Protocol=Cross-dataset2026.06 | 34.93 | — | 27.96 | — | — | — | 20.89 | — | 26.09 | |
| ASEN++Train Dataset=DARN, Evaluation Protocol=Cross-dataset2026.06 | 30.56 | — | 26.08 | — | — | — | 17.26 | — | 24.31 | |
| ASENTrain Dataset=DARN, Evaluation Protocol=Cross-dataset2026.06 | 29.36 | — | 25.08 | — | — | — | 16.86 | — | 23.35 | |
| ER-basedMethod category=CIL methods, Evaluation state=Final (Immediate) accuracy2026.03 | 18.23 | 28.66 | 21.74 | 27.59 | 25.79 | 23.62 | 14.92 | 25.09 | 22.08 | |
| Multi-head-basedMethod category=CIL methods, Evaluation state=Final (Immediate) accuracy2026.03 | 17.86 | 30.13 | 22.04 | 31.99 | 31.03 | 27.45 | 17.32 | 26.05 | 24.09 |