List your company
ResearchFormerly WizWand

Multimodal Model Evaluation on MME (Score)

1,636.5MME ScoreClean
-54.228384.711823.651,262.589Dec 11, 2024
Updated 4mo ago

Evaluation results

MethodLinks
CleanModel=Qwen2.5-VL, Vision Encoder=Qwen-ViT2024.121,636.5
CleanModel=LLaVA-1.5, Vision Encoder=CLIP-3362024.121,521.2
Img-EmbModel=LLaVA-1.5, Vision Encoder=CLIP-3362024.121,367.5
Both-EmbModel=LLaVA-1.5, Vision Encoder=CLIP-3362024.121,289.1
CleanModel=InstructBlip, Vision Encoder=EVA-CLIP2024.121,236.7
Text-EmbModel=LLaVA-1.5, Vision Encoder=CLIP-3362024.121,217.7
Both-EmbModel=InstructBlip, Vision Encoder=EVA-CLIP2024.121,115.9
Img-EmbModel=InstructBlip, Vision Encoder=EVA-CLIP2024.121,109.1
Img-EmbModel=Qwen2.5-VL, Vision Encoder=Qwen-ViT2024.12923.2
Text-EmbModel=InstructBlip, Vision Encoder=EVA-CLIP2024.12849.7
VEV-UAPModel=Qwen2.5-VL, Vision Encoder=Qwen-ViT2024.12758.2
VEV-UAPModel=InstructBlip, Vision Encoder=EVA-CLIP2024.12754.5
CleanModel=LLaVA, Vision Encoder=CLIP-3362024.12747.8
Both-EmbModel=LLaVA, Vision Encoder=CLIP-3362024.12699
Text-EmbModel=LLaVA, Vision Encoder=CLIP-3362024.12635.9
Img-EmbModel=LLaVA, Vision Encoder=CLIP-3362024.12621.5
VEV-UAPModel=LLaVA, Vision Encoder=CLIP-3362024.1223.6
VEV-UAPModel=LLaVA-1.5, Vision Encoder=CLIP-3362024.1210.8