EmoSpeech: Guiding FastSpeech2 Towards Emotional Text to Speech
This work addresses the need for more expressive and emotionally accurate speech synthesis, which is incremental as it builds upon the FastSpeech2 architecture with specific modifications.
The authors tackled the problem of synthesizing emotional speech in text-to-speech systems by modifying FastSpeech2, resulting in EmoSpeech, which achieved higher MOS scores and emotion recognition accuracy compared to existing models.
State-of-the-art speech synthesis models try to get as close as possible to the human voice. Hence, modelling emotions is an essential part of Text-To-Speech (TTS) research. In our work, we selected FastSpeech2 as the starting point and proposed a series of modifications for synthesizing emotional speech. According to automatic and human evaluation, our model, EmoSpeech, surpasses existing models regarding both MOS score and emotion recognition accuracy in generated speech. We provided a detailed ablation study for every extension to FastSpeech2 architecture that forms EmoSpeech. The uneven distribution of emotions in the text is crucial for better, synthesized speech and intonation perception. Our model includes a conditioning mechanism that effectively handles this issue by allowing emotions to contribute to each phone with varying intensity levels. The human assessment indicates that proposed modifications generate audio with higher MOS and emotional expressiveness.