This study reveals the importance of both visual and audio quality in omnidirectional videos, implying enhanced viewing experiences through integrated assessments.
Key Points
Comprehensive study integrates subjective and objective audio-visual quality assessments for ODVs.
A large-scale audio-visual quality assessment database named OAVQAD+ was established, containing 625 distorted sequences.
Development of benchmark models using support vector regression and multi-layer perceptron enhances OD-AVQA methods.
OmniAVNet effectively predicts the overall audio-visual quality, outperforming existing models on key databases.