Subcategory Recognition on FashionGen (test)
94.67AccuracyFAME-ViL
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| FAME-ViLtraining=Multi-Task Learning, backbone=ViT-B/162023.03 | 94.67 | 88.21 | 91.44 | |
| FAME-ViL(ST)training=Single-Task Learning, backbone=ViT-B/162023.03 | 94.33 | 86.21 | 90.27 | |
| FashionViL (vit)backbone=ViT-B/162023.03 | 94.01 | 85.77 | 89.89 | |
| MVLT2023.03 | 93.57 | 82.9 | 88.24 | |
| FashionViLbackbone=ResNet-502023.03 | 92.23 | 83.02 | 87.63 | |
| KaleidoBERT2023.03 | 88.07 | 63.6 | 75.84 | |
| FashionBERTOriginal Protocol=†2023.03 | 85.27 | 62 | 73.64 | |
| OSCAROriginal Protocol=†2023.03 | 84.23 | 59.1 | 71.67 |