Improving Speech Emotion Recognition with Unsupervised Speaking Style Transfer

no code implementations • 16 Nov 2022 • Leyuan Qu, Wei Wang, Cornelius Weber, Pengcheng Yue, Taihao Li, Stefan Wermter

Once training is completed, EmoAug enriches expressions of emotional speech with different prosodic attributes, such as stress, rhythm and intensity, by feeding different styles into the paralinguistic encoder.

Automatic Speech Recognition Automatic Speech Recognition (ASR) +4