Fine-grained Visual Recognition on Dog-120 (test)
85.8EMSigLIP-FT
Evaluation results
| Method | Links | ||
|---|---|---|---|
| SigLIP-FTModel Category=CLIP-based2026.04 | 85.8 | 93.4 | |
| HyMOR-GemmaModel Category=HyMOR framework2026.04 | 85.7 | 93.3 | |
| HyMOR-QwenModel Category=HyMOR framework2026.04 | 84.7 | 92.5 | |
| Finedefics-8BModel Category=Fine-grained MLLMs2026.04 | 79.8 | 91 | |
| SigLIPModel Category=CLIP-based2026.04 | 50.7 | 71.2 | |
| BioCLIPModel Category=CLIP-based2026.04 | 44.4 | 68 | |
| Qwen2.5-VL-3BModel Category=General MLLMs2026.04 | 26.7 | 63.6 | |
| Gemma-3-4BModel Category=General MLLMs2026.04 | 11.8 | 53.8 | |
| InternVL2.5-4BModel Category=General MLLMs2026.04 | 5 | 48 |