Image Classification on OxfordPets (Top-1 Accuracy)
92.6Top-1 AccuracyGPT-4V
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-4V2024.03 | 92.6 | |
| RARBase Model=Qwen-VL2024.03 | 88.1 | |
| RARBase Model=LLaVA1.52024.03 | 79.9 | |
| RARBase Model=Intern-IXC22024.03 | 79.3 | |
| RAR (LLaVA1.5)Shots=162024.03 | 72.7 | |
| CLIP+KNNShots=162024.03 | 68.3 | |
| RAR (LLaVA1.5)Shots=82024.03 | 67.7 | |
| LLaVA1.5 FinetuningShots=162024.03 | 62.4 | |
| CLIP+KNNShots=82024.03 | 61.8 | |
| RAR (LLaVA1.5)Shots=42024.03 | 60.9 | |
| CLIP+KNNShots=42024.03 | 55.6 | |
| RAR (LLaVA1.5)Shots=22024.03 | 54.7 | |
| CLIP+KNNShots=22024.03 | 49 | |
| RAR (LLaVA1.5)Shots=12024.03 | 47 | |
| LLaVA1.5 FinetuningShots=82024.03 | 46.1 | |
| CLIP+KNNShots=12024.03 | 41.3 | |
| LLaVA1.5 FinetuningShots=42024.03 | 28.9 | |
| LLaVA1.5 FinetuningShots=12024.03 | 16.3 | |
| LLaVA1.5 FinetuningShots=22024.03 | 16.3 |