List your company
ResearchFormerly WizWand

Visual Question Answering on VQAv2 (first quarter)

80ScoreClean
-2.88818.63140.1561.669Dec 11, 2024
Updated 4mo ago

Evaluation results

MethodLinks
CleanModel=LLaVA-1.5, Vision Encoder=CLIP-3362024.1280
CleanModel=Qwen2.5-VL, Vision Encoder=Qwen-ViT2024.1278.8
CleanModel=InstructBlip, Vision Encoder=EVA-CLIP2024.1275.6
Both-EmbModel=LLaVA-1.5, Vision Encoder=CLIP-3362024.1274.7
Img-EmbModel=LLaVA-1.5, Vision Encoder=CLIP-3362024.1272.1
Both-EmbModel=InstructBlip, Vision Encoder=EVA-CLIP2024.1270.4
Img-EmbModel=InstructBlip, Vision Encoder=EVA-CLIP2024.1269.9
Text-EmbModel=LLaVA-1.5, Vision Encoder=CLIP-3362024.1269.6
Text-EmbModel=InstructBlip, Vision Encoder=EVA-CLIP2024.1257.7
CleanModel=LLaVA, Vision Encoder=CLIP-3362024.1256.6
Both-EmbModel=LLaVA, Vision Encoder=CLIP-3362024.1252.8
Img-EmbModel=LLaVA, Vision Encoder=CLIP-3362024.1250.7
VEV-UAPModel=InstructBlip, Vision Encoder=EVA-CLIP2024.1246.8
Img-EmbModel=Qwen2.5-VL, Vision Encoder=Qwen-ViT2024.1245.7
Text-EmbModel=LLaVA, Vision Encoder=CLIP-3362024.1243.8
VEV-UAPModel=Qwen2.5-VL, Vision Encoder=Qwen-ViT2024.1242.8
VEV-UAPModel=LLaVA, Vision Encoder=CLIP-3362024.122.1
VEV-UAPModel=LLaVA-1.5, Vision Encoder=CLIP-3362024.120.3