DeepAI AI Chat
Log In Sign Up

Generative Speech Coding with Predictive Variance Regularization

by   W. Bastiaan Kleijn, et al.

The recent emergence of machine-learning based generative models for speech suggests a significant reduction in bit rate for speech codecs is possible. However, the performance of generative models deteriorates significantly with the distortions present in real-world input signals. We argue that this deterioration is due to the sensitivity of the maximum likelihood criterion to outliers and the ineffectiveness of modeling a sum of independent signals with a single autoregressive model. We introduce predictive-variance regularization to reduce the sensitivity to outliers, resulting in a significant increase in performance. We show that noise reduction to remove unwanted signals can significantly increase performance. We provide extensive subjective performance evaluations that show that our system based on generative modeling provides state-of-the-art coding performance at 3 kb/s for real-world speech signals at reasonable computational complexity.


Analysis by Adversarial Synthesis -- A Novel Approach for Speech Vocoding

Classical parametric speech coding techniques provide a compact represen...

A Streamwise GAN Vocoder for Wideband Speech Coding at Very Low Bit Rate

Recently, GAN vocoders have seen rapid progress in speech synthesis, sta...

Low Bit-Rate Wideband Speech Coding: A Deep Generative Model based Approach

Traditional low bit-rate speech coding approach only handles narrowband ...

Generative Models for Improved Naturalness, Intelligibility, and Voicing of Whispered Speech

This work adapts two recent architectures of generative models and evalu...

CoT: Cooperative Training for Generative Modeling

We propose Cooperative Training (CoT) for training generative models tha...

Learning to Adapt by Minimizing Discrepancy

We explore whether useful temporal neural generative models can be learn...

Varianceflow: High-Quality and Controllable Text-to-Speech using Variance Information via Normalizing Flow

There are two types of methods for non-autoregressive text-to-speech mod...