Next Tokens Denoising for Speech Synthesis

• URL: http://arxiv.org/abs/2507.22746v2
• Date: Fri, 01 Aug 2025 03:37:42 GMT
• Title: Next Tokens Denoising for Speech Synthesis
• Authors: Yanqing Liu, Ruiqing Xue, Chong Zhang, Yufei Liu, Gang Wang, Bohan Li, Yao Qian, Lei He, Shujie Liu, Sheng Zhao,
• Abstract summary: Dragon-FM is a novel text-to-speech (TTS) design that unifies AR and flow-matching.<n>It processes 48 kHz audio tokens in chunks at a compact rate of 12.5 tokens per second.<n>Experiments on podcast datasets demonstrate its capability to efficiently generate high-quality zero-shot podcasts.
• Score: 51.320443764269726
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: While diffusion and autoregressive (AR) models have significantly advanced generative modeling, they each present distinct limitations. AR models, which rely on causal attention, cannot exploit future context and suffer from slow generation speeds. Conversely, diffusion models struggle with key-value (KV) caching. To overcome these challenges, we introduce Dragon-FM, a novel text-to-speech (TTS) design that unifies AR and flow-matching. This model processes 48 kHz audio codec tokens in chunks at a compact rate of 12.5 tokens per second. This design enables AR modeling across chunks, ensuring global coherence, while parallel flow-matching within chunks facilitates fast iterative denoising. Thus, the model leverages KV-cache across chunks and utilizes bidirectional context within each chunk. Furthermore, it bridges continuous and discrete feature modeling, demonstrating that continuous AR flow-matching can predict discrete tokens with finite scalar quantizers. This efficient codec and fast chunk-autoregressive architecture also make the model highly effective for generating long-form content, such as podcasts. Experiments on podcast datasets demonstrate its capability to efficiently generate high-quality zero-shot podcasts.

Related papers

• Accelerating Diffusion Language Model Inference via Efficient KV Caching and Guided Diffusion [16.99620863197586]
  Diffusion language models offer parallel token generation and inherent bidirectionality.<n>State-of-the-art diffusion models (e.g., Dream 7B, LLaDA 8B) suffer from slow inference.<n>We introduce Guided Diffusion, a training-free method that uses a lightweight pretrained autoregressive model to supervise token unmasking.<n>For the first time, diffusion language models achieve a comparable and even faster latency as the widely adopted autoregressive models.
  arXiv  Detail & Related papers  (2025-05-27T17:39:39Z)
• Pseudo-Autoregressive Neural Codec Language Models for Efficient Zero-Shot Text-to-Speech Synthesis [64.12708207721276]
  We introduce a novel pseudo-autoregressive (PAR) language modeling approach that unifies AR and NAR modeling.<n>Building on PAR, we propose PALLE, a two-stage TTS system that leverages PAR for initial generation followed by NAR refinement.<n>Experiments demonstrate that PALLE, trained on LibriTTS, outperforms state-of-the-art systems trained on large-scale data.
  arXiv  Detail & Related papers  (2025-04-14T16:03:21Z)
• DiCoDe: Diffusion-Compressed Deep Tokens for Autoregressive Video Generation with Language Models [72.24305287508474]
  We introduce DiCoDe, a novel approach to generate videos with a language model in an autoregressive manner.<n>By treating videos as temporal sequences, DiCoDe fully harnesses the capabilities of language models for autoregressive generation.<n>We evaluate DiCoDe both quantitatively and qualitatively, demonstrating that it performs comparably to existing methods in terms of quality.
  arXiv  Detail & Related papers  (2024-12-05T18:57:06Z)
• ARLON: Boosting Diffusion Transformers with Autoregressive Models for Long Video Generation [83.62931466231898]
  This paper presents ARLON, a framework that boosts diffusion Transformers with autoregressive models for long video generation.<n>A latent Vector Quantized Variational Autoencoder (VQ-VAE) compresses the input latent space of the DiT model into compact visual tokens.<n>An adaptive norm-based semantic injection module integrates the coarse discrete visual units from the AR model into the DiT model.
  arXiv  Detail & Related papers  (2024-10-27T16:28:28Z)
• LANTERN: Accelerating Visual Autoregressive Models with Relaxed Speculative Decoding [30.630803933771865]
  Experimental results demonstrate the efficacy of our method in providing a substantial speed-up over speculative decoding.<n> LANTERN increases speed-ups by $mathbf1.75times$ and $mathbf1.82times$, as compared to greedy decoding and random sampling.
  arXiv  Detail & Related papers  (2024-10-04T12:21:03Z)
• Efficient Autoregressive Audio Modeling via Next-Scale Prediction [52.663934477127405]
  We analyze the token length of audio tokenization and propose a novel textbfScale-level textbfAudio textbfTokenizer (SAT)<n>Based on SAT, a scale-level textbfAcoustic textbfAutotextbfRegressive (AAR) modeling framework is proposed, which shifts the next-token AR prediction to next-scale AR prediction.
  arXiv  Detail & Related papers  (2024-08-16T21:48:53Z)
• Autoregressive Diffusion Transformer for Text-to-Speech Synthesis [39.32761051774537]
  We propose encoding audio as vector sequences in continuous space $mathbb Rd$ and autoregressively generating these sequences. High-bitrate continuous speech representation enables almost flawless reconstruction, allowing our model to achieve nearly perfect speech editing.
  arXiv  Detail & Related papers  (2024-06-08T18:57:13Z)
• From Discrete Tokens to High-Fidelity Audio Using Multi-Band Diffusion [84.138804145918]
  Deep generative models can generate high-fidelity audio conditioned on various types of representations. These models are prone to generate audible artifacts when the conditioning is flawed or imperfect. We propose a high-fidelity multi-band diffusion-based framework that generates any type of audio modality from low-bitrate discrete representations.
  arXiv  Detail & Related papers  (2023-08-02T22:14:29Z)
• Hierarchical Attention Encoder Decoder [2.4366811507669115]
  Autoregressive modeling can generate complex and novel sequences that have many real-world applications. These models must generate outputs autoregressively, which becomes time-consuming when dealing with long sequences. We propose a model based on the Hierarchical Recurrent Decoder architecture.
  arXiv  Detail & Related papers  (2023-06-01T18:17:23Z)
• FastLTS: Non-Autoregressive End-to-End Unconstrained Lip-to-Speech Synthesis [77.06890315052563]
  We propose FastLTS, a non-autoregressive end-to-end model which can directly synthesize high-quality speech audios from unconstrained talking videos with low latency. Experiments show that our model achieves $19.76times$ speedup for audio generation compared with the current autoregressive model on input sequences of 3 seconds.
  arXiv  Detail & Related papers  (2022-07-08T10:10:39Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.