His research interests mainly lie in machine learning, deep learning, and their applications on natural language and speech processing, including neural machine translation, pre-training, text to speech, automatic speech recognition, music generation, etc.