SDLGASAug 22, 2025

TaDiCodec: Text-aware Diffusion Speech Tokenizer for Speech Language Modeling

arXiv:2508.16790v19 citationsh-index: 13Has Code
Originality Highly original
AI Analysis

This addresses the problem of inefficient and complex speech tokenizers for researchers and developers in speech generation, offering a more streamlined and effective solution.

The paper tackles limitations in speech tokenizers for speech language models by introducing TaDiCodec, which uses a text-aware diffusion autoencoder for end-to-end optimization, achieving a low frame rate of 6.25 Hz and bitrate of 0.0875 kbps while maintaining superior performance on metrics like WER, SIM, and UMOS.

Speech tokenizers serve as foundational components for speech language models, yet current designs exhibit several limitations, including: 1) dependence on multi-layer residual vector quantization structures or high frame rates, 2) reliance on auxiliary pre-trained models for semantic distillation, and 3) requirements for complex two-stage training processes. In this work, we introduce the Text-aware Diffusion Transformer Speech Codec (TaDiCodec), a novel approach designed to overcome these challenges. TaDiCodec employs end-to-end optimization for quantization and reconstruction through a diffusion autoencoder, while integrating text guidance into the diffusion decoder to enhance reconstruction quality and achieve optimal compression. TaDiCodec achieves an extremely low frame rate of 6.25 Hz and a corresponding bitrate of 0.0875 kbps with a single-layer codebook for 24 kHz speech, while maintaining superior performance on critical speech generation evaluation metrics such as Word Error Rate (WER), speaker similarity (SIM), and speech quality (UTMOS). Notably, TaDiCodec employs a single-stage, end-to-end training paradigm, and obviating the need for auxiliary pre-trained models. We also validate the compatibility of TaDiCodec in language model based zero-shot text-to-speech with both autoregressive modeling and masked generative modeling, demonstrating its effectiveness and efficiency for speech language modeling, as well as a significantly small reconstruction-generation gap. We will open source our code and model checkpoints. Audio samples are are available at https:/tadicodec.github.io/. We release code and model checkpoints at https:/github.com/HeCheng0625/Diffusion-Speech-Tokenizer.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes