Share your thoughts, 1 month free Claude Pro on usSee more

Visual Question Answering on FVQA (test)

73.95Accuracy

MAIL

Updated 9d ago

Evaluation Results

Method
MAIL 2024.02	73.95	-	-	-	-	-
MIA 2026.04	69.6	-	-	-	-	-
HCNMN 2024.02	69.43	-	-	-	-	-
Gemini-3-Flash 2026.04	69.3	-	-	-	-	-
UnifER 2024.02	66.83	-	-	-	-	-
Memento 2026.04	66.3	-	-	-	-	-
Unsupervised MIA 2026.04	65.1	-	-	-	-	-
ReasoningBank 2026.04	64.7	-	-	-	-	-
MCAN 2024.02	64.47	-	-	-	-	-
ExpeL 2026.04	64.2	-	-	-	-	-
KI-Net 2024.02	63.78	-	-	-	-	-
XNM 2024.02	63.74	-	-	-	-	-
No Memory 2026.04	61.4	-	-	-	-	-
PMSR 2025.08	61.2	-	-	-	-	-
Deepeyes2 2026.04	60.6	-	-	-	-	-
DeepEyesV2 2025.08	60.6	-	-	-	-	-
RAG 2026.04	60.5	-	-	-	-	-
MMSearch-R1 2025.08	58.4	-	-	-	-	-
MMSearch-R1 2026.04	58	-	-	-	-	-
Mem0 2026.04	55.6	-	-	-	-	-
Qwen2.5-VL-32B+ReACT 2026.04	51.3	-	-	-	-	-
GPT-5.4 2026.04	50.8	-	-	-	-	-
GPT-4o 2026.04	41.7	-	-	-	-	-
A-Mem 2026.04	38.5	-	-	-	-	-
Gemini-2.5-Pro 2026.04	37.2	-	-	-	-	-
Qwen2.5-VL-7B+ReACT 2026.04	34.2	-	-	-	-	-
MMSearch-R1 2026.03	29.39	-	-	-	66.06	47.28
Qwen-VL-RAG 2026.03	24.83	-	-	-	100	100
Qwen2.5-VL-32B 2026.04	24.7	-	-	-	-	-
TA 2026.03	21.5	-	-	-	19.61	43.11
Qwen2.5-VL-7B 2026.04	20.9	-	-	-	-	-
OBM 2026.03	20.67	-	-	-	16.5	55.55
Qwen-VL-DA 2026.03	15.94	-	-	-	0	0
Qwen-VL-R1 2026.03	15.61	-	-	-	8.44	25.5
TIES 2026.03	15	-	-	-	1.94	32.72
DARE 2026.03	10.06	-	-	-	8.39	29.94
SVM-Question 2016.06	-	10.37	20.72	34.63	-	-
SVM-Image 2016.06	-	18.41	32.42	47.53	-	-
SVM-Question+Image 2016.06	-	18.89	32.78	48.13	-	-
LSTM-Question 2016.06	-	10.45	19.02	31.64	-	-
LSTM-Image 2016.06	-	20.55	36.01	55.74	-	-
LSTM-Question+Image 2016.06	-	22.97	36.76	54.19	-	-
LSTM-Question+Image+Pre-VQA 2016.06	-	24.98	40.4	57.27	-	-
Hie-Question+Image 2016.06	-	33.7	50	64.08	-	-
Hie-Question+Image+Pre-VQA 2016.06	-	43.14	59.44	72.2	-	-
FVQA model 2016.06	-	63.63	71.3	72.55	-	-
FVQA model 2016.06	-	52.56	59.72	60.58	-	-
FVQA model 2016.06	-	56.91	64.65	65.54	-	-
Ensemble 2016.06	-	58.76	-	-	-	-
Human 2016.06	-	77.99	-	-	-	-