Abstract: Teacher emotion expression is significant to classroom teaching. In this study, we used deep learning-based speech emotion recognition technology to construct a three-category emotion model ...

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.