Classification on AIR 10-way 0-shot
79.6AccuracySigLIP Zero-Shot
Evaluation Results
| Method | Links | |
|---|---|---|
| SigLIP Zero-ShotBackbone=SigLIP, Protocol=10-way 0-shot2026.03 | 79.6 | |
| Qwen2-VL HEC-TBackbone=DFN+LLM (Qwen2-VL), Protocol=10-way 0-shot2026.03 | 78.7 | |
| LLaVA-OV HEC-TBackbone=SigLIP+LLM (LLaVA-OV), Protocol=10-way 0-shot2026.03 | 73 | |
| DFN Zero-ShotBackbone=DFN, Protocol=10-way 0-shot2026.03 | 70.7 | |
| Qwen2-VL STBackbone=DFN+LLM (Qwen2-VL), Protocol=10-way 0-shot2026.03 | 66.4 | |
| LLaVA-OV STBackbone=SigLIP+LLM (LLaVA-OV), Protocol=10-way 0-shot2026.03 | 64.7 | |
| Qwen2-VL BaselineBackbone=DFN+LLM (Qwen2-VL), Protocol=10-way 0-shot2026.03 | 62.1 | |
| LLaVA-OV BaselineBackbone=SigLIP+LLM (LLaVA-OV), Protocol=10-way 0-shot2026.03 | 61.9 |