Image Classification on Vision-Few-Shot 12-dataset suite 4-shot
92.2PETS AccuracyHEC-V
Evaluation results
| Method | Links | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| HEC-VPrompt Conditioning=Domain prompt conditioning (DC), Hyperparameter Tuning=false, Few-shot setting=4-shot2026.03 | 92.2 | 78.8 | 85 | 82.4 | 95.5 | 71.8 | 62.2 | 85.3 | 98.5 | 89.8 | 72 | 75.5 | 82.4 | |
| Qwen2-VL Probing(DC)Model=Qwen2-VL, Backbone=DFN, Evaluation Protocol=Probing, Prompt Conditioning=Domain prompt conditioning (DC), Few-shot setting=4-shot2026.03 | 92 | 74 | 82.3 | 79.8 | 94.2 | 66.9 | 60.7 | 82.7 | 98.2 | 89.2 | 69.5 | 69.2 | 79.9 | |
| SAVsPrompt Conditioning=Domain prompt conditioning (DC), Hyperparameter Tuning=false, Few-shot setting=4-shot2026.03 | 91 | 72 | 80.5 | 81.7 | 94.4 | 70.5 | 59.3 | 84.9 | 97.8 | 89.5 | 69.8 | 68.1 | 80 | |
| DINOv3 ProbingBackbone=DINOv3, Evaluation Protocol=Probing, Few-shot setting=4-shot2026.03 | 88 | 77.2 | 82.8 | 72.7 | 95.4 | 63.6 | 56.9 | 74.3 | 99.5 | 79.4 | 77.3 | 53.1 | 76.7 | |
| Qwen2-VL Probing(TC)Model=Qwen2-VL, Backbone=DFN, Evaluation Protocol=Probing, Prompt Conditioning=Task prompt conditioning (TC), Few-shot setting=4-shot2026.03 | 84.9 | 75.6 | 81.4 | 77.2 | 93.4 | 58.2 | 40.9 | 81.1 | 98.1 | 79.8 | 65.1 | 71.3 | 75.6 | |
| DFN ProbingBackbone=DFN, Evaluation Protocol=Probing, Few-shot setting=4-shot2026.03 | 84.5 | 80.8 | 79.8 | 73.9 | 94.4 | 62.8 | 40 | 77.5 | 96.8 | 85.6 | 66.9 | 68.2 | 75.9 | |
| DINOv1 ProbingBackbone=DINOv1, Evaluation Protocol=Probing, Few-shot setting=4-shot2026.03 | 81.9 | 81.6 | 71.8 | 50.8 | 86.9 | 49.6 | 25.5 | 37.9 | 88.1 | 28.3 | 54.2 | 46.3 | 58.6 | |
| DINOv2 ProbingBackbone=DINOv2, Evaluation Protocol=Probing, Few-shot setting=4-shot2026.03 | 78.5 | 73.6 | 67.6 | 63.4 | 87 | 51.6 | 29.9 | 43 | 97 | 35.9 | 65.5 | 35.6 | 60.7 | |
| OpenCLIP ProbingBackbone=OpenCLIP, Evaluation Protocol=Probing, Few-shot setting=4-shot2026.03 | 73 | 75.6 | 79.5 | 72.1 | 90.4 | 60.3 | 29.6 | 62.9 | 89 | 75 | 51.2 | 62.7 | 68.4 | |
| OpenAI CLIP ProbingBackbone=OpenAI CLIP, Evaluation Protocol=Probing, Few-shot setting=4-shot2026.03 | 72.9 | 73.6 | 81.5 | 73.2 | 90.2 | 55.8 | 28.2 | 73.4 | 89.5 | 56.9 | 53.6 | 56.5 | 67.1 |