List your company
ResearchFormerly WizWand

Visual Reasoning on VizWiz

0.11ECEPlausibility
0.094640.198320.3020.40568Sep 30, 2025
Updated 6mo ago

Evaluation results

MethodLinks
PlausibilityModel=LLaVA2025.090.11
Avg(VF, Contr.)Model=Qwen2.52025.090.138
Visual FidelityModel=Qwen2.52025.090.139
Avg(VF, Contr.)Model=LLaVA2025.090.139
Visual FidelityModel=GPT-4o2025.090.16
Avg(VF, Contr.)Model=Avg2025.090.164
InformativenessModel=Qwen2.52025.090.168
Visual FidelityModel=Avg2025.090.19
Avg(VF, Contr.)Model=GPT-4o2025.090.216
PlausibilityModel=Avg2025.090.248
Min(VF, Contr.)Model=LLaVA2025.090.265
Visual FidelityModel=LLaVA2025.090.271
Min(VF, Contr.)Model=Qwen2.52025.090.284
Prod(VF, Contr.)Model=LLaVA2025.090.29
SimulatabilityModel=Qwen2.52025.090.292
ContrastivenessModel=Qwen2.52025.090.292
ContrastivenessModel=LLaVA2025.090.294
Prod(VF, Contr.)Model=Qwen2.52025.090.296
SimulatabilityModel=LLaVA2025.090.305
PlausibilityModel=Qwen2.52025.090.31
PlausibilityModel=GPT-4o2025.090.325
Min(VF, Contr.)Model=Avg2025.090.325
ContrastivenessModel=Avg2025.090.337
Prod(VF, Contr.)Model=Avg2025.090.337
InformativenessModel=Avg2025.090.348
SimulatabilityModel=Avg2025.090.364
ContrastivenessModel=GPT-4o2025.090.424
Min(VF, Contr.)Model=GPT-4o2025.090.425
Prod(VF, Contr.)Model=GPT-4o2025.090.426
InformativenessModel=LLaVA2025.090.429
InformativenessModel=GPT-4o2025.090.446
SimulatabilityModel=GPT-4o2025.090.494