AutoTTS: End-to-End Text-to-Speech Synthesis through Differentiable Duration Modeling

TL;DR

AutoTTS introduces a differentiable duration modeling approach for end-to-end text-to-speech synthesis, enabling high-quality speech generation with a simpler training process by jointly learning alignments and waveform generation.

Contribution

The paper presents a novel differentiable duration method that allows joint training of alignment and synthesis in an end-to-end TTS model, simplifying the pipeline.

Findings

Achieves competitive speech quality with a simpler training pipeline

Uses a soft-duration mechanism for monotonic alignment learning

Combines adversarial training with duration matching for high fidelity

Abstract

Parallel text-to-speech (TTS) models have recently enabled fast and highly-natural speech synthesis. However, they typically require external alignment models, which are not necessarily optimized for the decoder as they are not jointly trained. In this paper, we propose a differentiable duration method for learning monotonic alignments between input and output sequences. Our method is based on a soft-duration mechanism that optimizes a stochastic process in expectation. Using this differentiable duration method, we introduce AutoTTS, a direct text-to-waveform speech synthesis model. AutoTTS enables high-fidelity speech synthesis through a combination of adversarial training and matching the total ground-truth duration. Experimental results show that our model obtains competitive results while enjoying a much simpler training pipeline. Audio samples are available online.