Audio-Visual Question Answering
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
85.09AV Localis Accuracy
96
Feb 26, 2026
80.7Acc (Avg)
94
Jul 7, 2026
87.3Accuracy
85
Jun 2, 2026
94.3Total Accuracy
49
Jun 8, 2026
85.68Audio QA Count (Head)
41
Jul 7, 2026
81.3Accuracy
38
Jul 7, 2026
78.4Total Accuracy
38
Jul 7, 2026
75.44Total Accuracy
38
Jul 7, 2026
85.15Music-AVQA Clean Accuracy
25
Apr 7, 2026
29.1Miss Rate
18
Feb 26, 2026
0.356Accuracy
18
Feb 26, 2026
50Accuracy
18
Feb 26, 2026
165.1CIDEr
15
May 27, 2026
83.38Audio Count Acc
14
Mar 9, 2026
54.68Accuracy (V+A+L)
12
Mar 24, 2026
44.67M.J. Score
12
Feb 26, 2026
81.58Accuracy (MUSIC-AVQA)
10
May 12, 2026
69.1Average Score
9
May 27, 2026
88.24Existence Accuracy
9
Feb 26, 2026
70.9Metric 30
8
May 26, 2026
64.4SR
8
May 26, 2026
73.6Score
8
Mar 17, 2026
59.9Score
8
Mar 17, 2026
75Score
8
Mar 17, 2026
83.62Existential Score
8
Feb 26, 2026