Speech Sentiment Analysis via Pre-trained Features from End-to-end ASR Models
This work addresses sentiment analysis from speech for applications like human-computer interaction, but it is incremental as it adapts existing ASR features to a new task.
The paper tackles speech sentiment analysis by using pre-trained features from end-to-end ASR models, achieving an accuracy improvement from 66.6% to 71.7% on IEMOCAP and 70.10% on a new large-scale dataset.
In this paper, we propose to use pre-trained features from end-to-end ASR models to solve speech sentiment analysis as a down-stream task. We show that end-to-end ASR features, which integrate both acoustic and text information from speech, achieve promising results. We use RNN with self-attention as the sentiment classifier, which also provides an easy visualization through attention weights to help interpret model predictions. We use well benchmarked IEMOCAP dataset and a new large-scale speech sentiment dataset SWBD-sentiment for evaluation. Our approach improves the-state-of-the-art accuracy on IEMOCAP from 66.6% to 71.7%, and achieves an accuracy of 70.10% on SWBD-sentiment with more than 49,500 utterances.