Prosody-TTS: An end-to-end speech synthesis system with prosody control

TL;DR

Prosody-TTS is an end-to-end speech synthesis system that allows explicit prosody control and achieves state-of-the-art naturalness with low inference time, combining statistical and neural methods.

Contribution

It introduces a novel end-to-end model that explicitly controls prosody by modeling phoneme duration and f0, improving naturalness and efficiency.

Findings

Achieves a mean opinion score of 4.08 on Telugu data.

Provides explicit prosody control via f0 and duration modeling.

Demonstrates low inference time and state-of-the-art performance.

Abstract

End-to-end text-to-speech synthesis systems achieved immense success in recent times, with improved naturalness and intelligibility. However, the end-to-end models, which primarily depend on the attention-based alignment, do not offer an explicit provision to modify/incorporate the desired prosody while synthesizing the signal. Moreover, the state-of-the-art end-to-end systems use autoregressive models for synthesis, making the prediction sequential. Hence, the inference time and the computational complexity are quite high. This paper proposes Prosody-TTS, an end-to-end speech synthesis model that combines the advantages of statistical parametric models and end-to-end neural network models. It also has a provision to modify or incorporate the desired prosody by controlling the fundamental frequency (f0) and the phone duration. Generating speech samples with appropriate prosody and rhythm helps in improving the naturalness of the synthesized speech. We explicitly model the duration of the phoneme and the f0 to have control over them during the synthesis. The model is trained in an end-to-end fashion to directly generate the speech waveform from the input text, which in turn depends on the auxiliary subtasks of predicting the phoneme duration, f0, and mel spectrogram. Experiments on the Telugu language data of the IndicTTS database show that the proposed Prosody-TTS model achieves state-of-the-art performance with a mean opinion score of 4.08, with a very low inference time.