MCM-AVQA improves correlation with human quality scores by using modality-specific confidence to suppress unreliable signals during audio-visual fusion under asymmetric distortions.
The naive late-fusion baseline (A VM-, VCM-, ACM-) has PLCC/SROCC values of 0.907/0.894 on UnB-A VQ and 0.916/0.896 on LIVE-SJTU
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.MM 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Multimodal Confidence Modeling in Audio-Visual Quality Assessment
MCM-AVQA improves correlation with human quality scores by using modality-specific confidence to suppress unreliable signals during audio-visual fusion under asymmetric distortions.