SoCodec: A Semantic-Ordered Multi-Stream Speech Codec for Efficient Language Model Based Text-to-Speech Synthesis

TL;DR

SoCodec introduces a semantic-ordered multi-stream speech codec that compresses speech into shorter, ordered semantic sequences, enabling more efficient and effective language model-based text-to-speech synthesis with improved performance.

Contribution

It proposes a novel semantic-ordered multi-stream speech codec with ordered product quantization, enhancing compression and autoregressive generation in TTS systems.

Findings

Achieves 12x frame compression from 20ms to 240ms.

Outperforms baseline systems in TTS quality.

Validates importance of ordered multi-stream semantic representation.

Abstract

The long speech sequence has been troubling language models (LM) based TTS approaches in terms of modeling complexity and efficiency. This work proposes SoCodec, a semantic-ordered multi-stream speech codec, to address this issue. It compresses speech into a shorter, multi-stream discrete semantic sequence with multiple tokens at each frame. Meanwhile, the ordered product quantization is proposed to constrain this sequence into an ordered representation. It can be applied with a multi-stream delayed LM to achieve better autoregressive generation along both time and stream axes in TTS. The experimental result strongly demonstrates the effectiveness of the proposed approach, achieving superior performance over baseline systems even if compressing the frameshift of speech from 20ms to 240ms (12x). The ablation studies further validate the importance of learning the proposed ordered multi-stream semantic representation in pursuing shorter speech sequences for efficient LM-based TTS.