Multi-label Classification on ImageNet Seg
89.4Top-1 Accuracy+Multi-label FT
Evaluation results
| Method | Links | ||
|---|---|---|---|
| +Multi-label FTModel=ViT-large2026.03 | 89.4 | 91.9 | |
| Multi-label E2EModel=ViT-large2026.03 | 89.4 | 91.9 | |
| +IN1k-Mul FTBackbone=ViT-B/16, Resolution=224, Pre-training=ImageNet-21K (MIIL), Supervision=Multi-label (Mul), Training Protocol=Fine-tuning2026.03 | 89.2 | 91.8 | |
| Multi-label E2EModel=ViT-base2026.03 | 89.1 | 91.5 | |
| IN1k-Mul E2EBackbone=ViT-B/16, Resolution=224, Supervision=Multi-label (Mul), Training Protocol=End-to-end2026.03 | 89.1 | 91.5 | |
| +Multi-label FTModel=ViT-base2026.03 | 88.8 | 91.3 | |
| +IN1k-Sig FTBackbone=ViT-B/16, Resolution=224, Pre-training=ImageNet-21K (MIIL), Supervision=Single-label (Sig), Training Protocol=Fine-tuning2026.03 | 88.8 | 91.4 | |
| Single-label E2EModel=ViT-large2026.03 | 88.5 | 91.2 | |
| Multi-label E2EModel=ViT-small2026.03 | 88.3 | 91.1 | |
| Single-label E2EModel=ViT-base2026.03 | 88.2 | 90.9 | |
| +Multi-label FTModel=ViT-small2026.03 | 87.8 | 90.9 | |
| Single-label E2EModel=ViT-small2026.03 | 87.2 | 89.7 | |
| +Multi-label FTModel=ResNet-1012026.03 | 87.1 | 90.1 | |
| Multi-label E2EModel=ResNet-1012026.03 | 86.9 | 89.9 | |
| +Multi-label FTModel=ResNet-502026.03 | 85.9 | 89.1 | |
| Multi-label E2EModel=ResNet-502026.03 | 85.5 | 88.8 | |
| Single-label E2EModel=ResNet-1012026.03 | 85.3 | 88.2 | |
| Single-label E2EModel=ResNet-502026.03 | 84.4 | 87.7 |