Image Classification on CIFAR-100-LT (Consistent split)
82.2AccuracyFixMatch + AdaptFormer + DECON
Evaluation results
| Method | Links | |
|---|---|---|
| FixMatch + AdaptFormer + DECONFine-tuning strategy=AdaptFormer, Pre-trained model=CLIP2024.06 | 82.2 | |
| FixMatch + AdaptFormer + ACRFine-tuning strategy=AdaptFormer, Pre-trained model=CLIP2024.06 | 81.7 | |
| FixMatch + VPT + DECONFine-tuning strategy=Visual Prompt Tuning (VPT), Pre-trained model=CLIP2024.06 | 81.3 | |
| FixMatch + VPT + ACRFine-tuning strategy=Visual Prompt Tuning (VPT), Pre-trained model=CLIP2024.06 | 80.8 | |
| FixMatch + AdaptFormerFine-tuning strategy=AdaptFormer, Pre-trained model=CLIP2024.06 | 79.7 | |
| FixMatch + VPTFine-tuning strategy=Visual Prompt Tuning (VPT), Pre-trained model=CLIP2024.06 | 78.6 |