Autoregressive Diffusion Transformer for Text-to-Speech Synthesis

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Liu, Zhijun, Wang, Shuai, Inoue, Sho, Bai, Qibing, Li, Haizhou
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914830025228288
author Liu, Zhijun
Wang, Shuai
Inoue, Sho
Bai, Qibing
Li, Haizhou
author_facet Liu, Zhijun
Wang, Shuai
Inoue, Sho
Bai, Qibing
Li, Haizhou
contents Audio language models have recently emerged as a promising approach for various audio generation tasks, relying on audio tokenizers to encode waveforms into sequences of discrete symbols. Audio tokenization often poses a necessary compromise between code bitrate and reconstruction accuracy. When dealing with low-bitrate audio codes, language models are constrained to process only a subset of the information embedded in the audio, which in turn restricts their generative capabilities. To circumvent these issues, we propose encoding audio as vector sequences in continuous space $\mathbb R^d$ and autoregressively generating these sequences using a decoder-only diffusion transformer (ARDiT). Our findings indicate that ARDiT excels in zero-shot text-to-speech and exhibits performance that compares to or even surpasses that of state-of-the-art models. High-bitrate continuous speech representation enables almost flawless reconstruction, allowing our model to achieve nearly perfect speech editing. Our experiments reveal that employing Integral Kullback-Leibler (IKL) divergence for distillation at each autoregressive step significantly boosts the perceived quality of the samples. Simultaneously, it condenses the iterative sampling process of the diffusion model into a single step. Furthermore, ARDiT can be trained to predict several continuous vectors in one step, significantly reducing latency during sampling. Impressively, one of our models can generate $170$ ms of $24$ kHz speech per evaluation step with minimal degradation in performance. Audio samples are available at http://ardit-tts.github.io/ .
format Preprint
id arxiv_https___arxiv_org_abs_2406_05551
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Autoregressive Diffusion Transformer for Text-to-Speech Synthesis
Liu, Zhijun
Wang, Shuai
Inoue, Sho
Bai, Qibing
Li, Haizhou
Audio and Speech Processing
Artificial Intelligence
Computation and Language
Machine Learning
Sound
Audio language models have recently emerged as a promising approach for various audio generation tasks, relying on audio tokenizers to encode waveforms into sequences of discrete symbols. Audio tokenization often poses a necessary compromise between code bitrate and reconstruction accuracy. When dealing with low-bitrate audio codes, language models are constrained to process only a subset of the information embedded in the audio, which in turn restricts their generative capabilities. To circumvent these issues, we propose encoding audio as vector sequences in continuous space $\mathbb R^d$ and autoregressively generating these sequences using a decoder-only diffusion transformer (ARDiT). Our findings indicate that ARDiT excels in zero-shot text-to-speech and exhibits performance that compares to or even surpasses that of state-of-the-art models. High-bitrate continuous speech representation enables almost flawless reconstruction, allowing our model to achieve nearly perfect speech editing. Our experiments reveal that employing Integral Kullback-Leibler (IKL) divergence for distillation at each autoregressive step significantly boosts the perceived quality of the samples. Simultaneously, it condenses the iterative sampling process of the diffusion model into a single step. Furthermore, ARDiT can be trained to predict several continuous vectors in one step, significantly reducing latency during sampling. Impressively, one of our models can generate $170$ ms of $24$ kHz speech per evaluation step with minimal degradation in performance. Audio samples are available at http://ardit-tts.github.io/ .
title Autoregressive Diffusion Transformer for Text-to-Speech Synthesis
topic Audio and Speech Processing
Artificial Intelligence
Computation and Language
Machine Learning
Sound
url https://arxiv.org/abs/2406.05551