A Methodology for Controlling the Emotional Expressiveness in Synthetic Speech -- a Deep Learning approach
This work addresses the need for more expressive and controllable synthetic speech in applications like human-computer interaction, though it appears incremental by building on existing TTS methods.
The researchers tackled the problem of controlling emotional expressiveness in synthetic speech by developing a Text-to-Speech system that integrates emotional data collection, automatic annotation, and deep learning-based generation, resulting in improved intelligibility and emotion perception through fine-tuning from neutral to emotional TTS.
In this project, we aim to build a Text-to-Speech system able to produce speech with a controllable emotional expressiveness. We propose a methodology for solving this problem in three main steps. The first is the collection of emotional speech data. We discuss the various formats of existing datasets and their usability in speech generation. The second step is the development of a system to automatically annotate data with emotion/expressiveness features. We compare several techniques using transfer learning to extract such a representation through other tasks and propose a method to visualize and interpret the correlation between vocal and emotional features. The third step is the development of a deep learning-based system taking text and emotion/expressiveness as input and producing speech as output. We study the impact of fine tuning from a neutral TTS towards an emotional TTS in terms of intelligibility and perception of the emotion.