#Dasheng

Researchers have developed a flexible speech enhancement system using pre-trained generative audioencoders and vocoders, achieving superior speech quality and efficiency compared to existing models.