Abstract: Noise robustness is critical when applying automatic speech recognition (ASR) in real-world scenarios. One solution involves using speech enhancement (SE) models as the front end of ASR.
🌐 24+ Languages: Arabic, Chinese, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Norwegian, Polish ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results