A Robust, Explainable, and Efficient Deep Metric for Speech and General Audio Quality | ArxivCSExplorer