UFANS: U-shaped Fully-Parallel Acoustic Neural Structure For Statistical Parametric Speech Synthesis With 20X Faster

Abstract

Neural networks with Auto-regressive structures, such as Recurrent Neural Networks (RNNs), have become the most appealing structures for acoustic modeling of parametric text to speech synthesis (TTS) in ecent studies. Despite the prominent capacity to capture long-term dependency, these models consist of massive sequential computations that cannot be fully parallel. In this paper, we propose a U-shaped Fully-parallel Acoustic Neural Structure (UFANS), which is a deconvolutional alternative of RNNs for Statistical Parametric Speech Synthesis (SPSS). The experiments verify that our proposed model is over 20 times faster than RNN based acoustic model, both training and inference on GPU with comparable speech quality. Furthermore, We also investigate that how long information dependence really matters to synthesized speech quality.

References in corpus (5)

• Adam: A Method for Stochastic Optimization
• WaveNet: A Generative Model for Raw Audio
• Opening the Black Box of Deep Neural Networks via Information
• Efficiently Trainable Text-to-Speech System Based on Deep Convolutional Networks with Guided Attention
• Boosting Noise Robustness of Acoustic Model via Deep Adversarial Training

Cited by in corpus (1)

• FPETS : Fully Parallel End-to-End Text-to-Speech System