Classification on UCF101 zero-shot
94.2ACCLLaVA-OV HEC-T
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| LLaVA-OV HEC-TBackbone=SigLIP+LLM (LLaVA-OV), Protocol=10-way 0-shot2026.03 | 94.2 | — | — | |
| Qwen2-VL HEC-TBackbone=DFN+LLM (Qwen2-VL), Protocol=10-way 0-shot2026.03 | 92.6 | — | — | |
| LLaVA-OV BaselineBackbone=SigLIP+LLM (LLaVA-OV), Protocol=10-way 0-shot2026.03 | 92.1 | — | — | |
| LLaVA-OV STBackbone=SigLIP+LLM (LLaVA-OV), Protocol=10-way 0-shot2026.03 | 91.7 | — | — | |
| Qwen2-VL STBackbone=DFN+LLM (Qwen2-VL), Protocol=10-way 0-shot2026.03 | 90.5 | — | — | |
| DFN Zero-ShotBackbone=DFN, Protocol=10-way 0-shot2026.03 | 87.8 | — | — | |
| SigLIP Zero-ShotBackbone=SigLIP, Protocol=10-way 0-shot2026.03 | 87.1 | — | — | |
| Qwen2-VL BaselineBackbone=DFN+LLM (Qwen2-VL), Protocol=10-way 0-shot2026.03 | 85.6 | — | — | |
| TTABackbone=SigLIP-Base, Zero-shot=true2024.12 | 60.14 | 1.88 | 13.47 | |
| BayesVLMBackbone=SigLIP-Base, Zero-shot=true2024.12 | 59.69 | 1.82 | 9.6 | |
| SigLIPBackbone=SigLIP-Base, Zero-shot=true2024.12 | 59.61 | 1.9 | 12.72 | |
| SigLIP (temp. scaling)Backbone=SigLIP-Base, Zero-shot=true2024.12 | 59.61 | 1.77 | 6.69 |