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

Text-based Visual Question Answering on TextVQA (TQA)

66.6Score

VILA

24.27235.26146.2557.239Feb 15, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.02
66.6
2026.02
64.9
2026.02
64.7
2026.02
58.2
2026.02
57
2026.02
56.8
2026.02
54.1
2026.02
53.7
2026.02
51.6
2026.02
50.1
2026.02
48.3
2026.02
47.7
2026.02
46.1
2026.02
25.9