SoundStream: An End-to-End Neural Audio Codec

07/07/2021
by   Neil Zeghidour, et al.
6

We present SoundStream, a novel neural audio codec that can efficiently compress speech, music and general audio at bitrates normally targeted by speech-tailored codecs. SoundStream relies on a model architecture composed by a fully convolutional encoder/decoder network and a residual vector quantizer, which are trained jointly end-to-end. Training leverages recent advances in text-to-speech and speech enhancement, which combine adversarial and reconstruction losses to allow the generation of high-quality audio content from quantized embeddings. By training with structured dropout applied to quantizer layers, a single model can operate across variable bitrates from 3kbps to 18kbps, with a negligible quality loss when compared with models trained at fixed bitrates. In addition, the model is amenable to a low latency implementation, which supports streamable inference and runs in real time on a smartphone CPU. In subjective evaluations using audio at 24kHz sampling rate, SoundStream at 3kbps outperforms Opus at 12kbps and approaches EVS at 9.6kbps. Moreover, we are able to perform joint compression and enhancement either at the encoder or at the decoder side with no additional latency, which we demonstrate through background noise suppression for speech.

READ FULL TEXT

page 2

page 4

page 5

research
09/08/2023

A Two-Stage Training Framework for Joint Speech Compression and Enhancement

This paper considers the joint compression and enhancement problem for s...
research
10/24/2022

High Fidelity Neural Audio Compression

We introduce a state-of-the-art real-time, high-fidelity, audio codec le...
research
08/08/2021

Audio Spectral Enhancement: Leveraging Autoencoders for Low Latency Reconstruction of Long, Lossy Audio Sequences

With active research in audio compression techniques yielding substantia...
research
04/01/2021

Multi-rate attention architecture for fast streamable Text-to-speech spectrum modeling

Typical high quality text-to-speech (TTS) systems today use a two-stage ...
research
03/10/2023

An End-to-End Neural Network for Image-to-Audio Transformation

This paper describes an end-to-end (E2E) neural architecture for the aud...
research
08/17/2020

Efficient Low-Latency Speech Enhancement with Mobile Audio Streaming Networks

We propose Mobile Audio Streaming Networks (MASnet) for efficient low-la...
research
10/05/2021

Late reverberation suppression using U-nets

In real-world settings, speech signals are almost always affected by rev...

Please sign up or login with your details

Forgot password? Click here to reset