LPCNet: Improving Neural Speech Synthesis Through Linear Prediction

TL;DR

LPCNet is a neural speech synthesis model that combines linear prediction with recurrent neural networks to achieve high-quality speech with lower computational complexity, enabling deployment on resource-constrained devices.

Contribution

The paper introduces LPCNet, a novel neural speech synthesis model that significantly improves efficiency while maintaining high speech quality, compared to existing models like WaveRNN.

Findings

LPCNet outperforms WaveRNN in speech quality at the same network size.

LPCNet achieves high-quality synthesis with less than 3 GFLOPS complexity.

LPCNet enables real-time speech synthesis on low-power devices.

Abstract

Neural speech synthesis models have recently demonstrated the ability to synthesize high quality speech for text-to-speech and compression applications. These new models often require powerful GPUs to achieve real-time operation, so being able to reduce their complexity would open the way for many new applications. We propose LPCNet, a WaveRNN variant that combines linear prediction with recurrent neural networks to significantly improve the efficiency of speech synthesis. We demonstrate that LPCNet can achieve significantly higher quality than WaveRNN for the same network size and that high quality LPCNet speech synthesis is achievable with a complexity under 3 GFLOPS. This makes it easier to deploy neural synthesis applications on lower-power devices, such as embedded systems and mobile phones.