List your company

Visual Question Answering on A-OKVQA Open-Ended

72.14AccuracyCVLM
52.972857.948962.92567.9011Feb 21, 2024

Evaluation results

MethodLinks
CVLMLLM=Qwen-VL2024.0272.14
CVLM (3M IKPairs, Objects=3)LLM=Vicuna-7B, Objects=32024.0270.91
CVLM (3M IKPairs, Objects=1)LLM=Vicuna-7B, Objects=12024.0270.48
CVLM (3M IKPairs, Objects=5)LLM=Vicuna-7B, Objects=52024.0270.39
CVLM (3M IKPairs) w/o FKALLM=Vicuna-7B2024.0270.22
CVLM (3M IKPairs, Objects=8)LLM=Vicuna-7B, Objects=82024.0269.08
InstructBLIPLLM=Vicuna-7B2024.0267.86
Qwen-VLLLM=Qwen2024.0267.25
CVLMLLM=Vicuna-7B2024.0266.99
CVLM w/o FKALLM=Vicuna-7B2024.0265.59
LLaVA-v1.5†LLM=Vicuna-7B2024.0265.27
CVLM w/o (FKA & VKA)LLM=Vicuna-7B2024.0264.37
InstructBLIPLLM=Flan-T5-XXL2024.0262.01
InstructBLIPLLM=Flan-T5-XL2024.0261.05
BLIP2LLM=Flan-T5-XXL2024.0253.71