Researchers at The Ohio State University have developed an AI model that improves audio quality by combining subjective ratings of sound quality with a speech enhancement algorithm.
The model outperformed other approaches in minimizing noisy audio and the predicted quality scores were strongly correlated to human judgments.