Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

TextVQA

Benchmarks

Task NameDataset NameSOTA ResultTrend
Visual Question AnsweringTextVQA
Accuracy85.4
1,117
Text-based Visual Question AnsweringTextVQA
Accuracy86.2
496
Visual Question AnsweringTextVQA (val)
VQA Score7,040
309
Text-based Visual Question AnsweringTextVQA (val)
Accuracy85.5
146
Visual Question AnsweringTextVQA (test)
Accuracy81.1
124
Visual Question AnsweringTextVQA
Accuracy97.15
79
Visual Question AnsweringTextVQA
Accuracy88.7
69
Visual Question AnsweringTextVQA v1.0 (val)
Accuracy85.5
69
Text-based Visual Question AnsweringTextVQA (VQA^T)
Accuracy70.4
65
Text-based Visual Question AnsweringTextVQA
Score63.2
38
Visual Question AnsweringTextVQA
Clean Accuracy70.3
37
Visual Question AnsweringTextVQA
VQA Accuracy39
33
Visual Question AnsweringTextVQA v1.0 (test)
Accuracy86.79
27
Visual Question AnsweringTextVQA
Exact Match (EM)82.74
23
Visual Question AnsweringTextVQA 130 (val)
Score86.5
23
Text-based Visual Question AnsweringTextVQA 52
Accuracy63.8
23
OCR-related Understanding TasksTextVQA (val)
Accuracy86.62
22
Text-based Visual Question AnsweringTextVQA
Average Score100
21
Image UnderstandingTextVQA
Accuracy85.76
16
Visual Question AnsweringTextVQA 1k (test)
ASR (%)96.46
15
Text-based Visual Question AnsweringTextVQA (TQA)
Score66.6
14
Copyright trackingTextVQA
ASR47
13
OCR-based Visual Question AnsweringTextVQA 2019 (val)
Accuracy83.8
13
Visual Question AnsweringTextVQA
Score69.89
12
OCR VQATextVQA (test)
Pre Accuracy61.9
10
Showing 25 of 48 rows