Long-tailed Image Recognition on Places-LT (test)
31.4Accuracy (Old)SimGCD
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| SimGCDBackbone=ViT-B/16, Pre-trained=DINO, Fine-tuning protocol=fine-tune last block2023.10 | 31.4 | 18.2 | 24.8 | |
| BaCon-SBackbone=ViT-B/16, Pre-trained=DINO, Fine-tuning protocol=fine-tune last block, pseudo-labeling branch=SimGCD-style2023.10 | 31.1 | 28.4 | 29.9 | |
| BaCon-OBackbone=ViT-B/16, Pre-trained=DINO, Fine-tuning protocol=fine-tune last block, pseudo-labeling branch=ORCA-style2023.10 | 30.7 | 25.6 | 28.1 | |
| OpenConBackbone=ViT-B/16, Pre-trained=DINO, Fine-tuning protocol=fine-tune last block2023.10 | 30.6 | 12.4 | 21.6 | |
| GCDBackbone=ViT-B/16, Pre-trained=DINO, Fine-tuning protocol=fine-tune last block2023.10 | 29.8 | 22.7 | 26.2 | |
| ORCABackbone=ViT-B/16, Pre-trained=DINO, Fine-tuning protocol=fine-tune last block2023.10 | 21.5 | 6.9 | 14.2 |