Multimodal Understanding on MMMU (Corruption Metrics: Noise, Blur, Weather, Digital)
37.3Accuracy (Noise)Latent Denoising
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Latent DenoisingBackbone=LLaVA+CLIP, Corruption Severity=52026.04 | 37.3 | 35.5 | 36.7 | 36.9 | |
| BaselineBackbone=LLaVA+CLIP, Corruption Severity=52026.04 | 35.5 | 34.6 | 33.6 | 35.2 |