Get our free extension to see links to code for papers anywhere online!

Chrome logo  Add to Chrome

Firefox logo Add to Firefox


Sentiment-Aware Automatic Speech Recognition pre-training for enhanced Speech Emotion Recognition

Jan 27, 2022
Ayoub Ghriss, Bo Yang, Viktor Rozgic, Elizabeth Shriberg, Chao Wang


Share this with someone who'll enjoy it:


We propose a novel multi-task pre-training method for Speech Emotion Recognition (SER). We pre-train SER model simultaneously on Automatic Speech Recognition (ASR) and sentiment classification tasks to make the acoustic ASR model more ``emotion aware''. We generate targets for the sentiment classification using text-to-sentiment model trained on publicly available data. Finally, we fine-tune the acoustic ASR on emotion annotated speech data. We evaluated the proposed approach on the MSP-Podcast dataset, where we achieved the best reported concordance correlation coefficient (CCC) of 0.41 for valence prediction.

* ICASSP 2022 


   Access Paper Source



Share this with someone who'll enjoy it: