mlsanityData to 14 Oct 2026

WaveNeXt 2: ConvNeXt-Based Fast Neural Vocoders With Residual Denoising and Sub-Modeling for GAN and Diffusion Models

arXiv 2605.25506 ↗First posted 25 May 2026eess.AScited by 0

On the map: Diffusion models → Audio, speech and music

Abstract (arXiv)

Most neural vocoders are limited to one type: either GAN or diffusion-based. While state-of-the-art models like Vocos and WaveNeXt use powerful ConvNeXt-based generators, they have only been used in GAN frameworks and have limited performance in multi-speaker settings. Moreover, diffusion models, despite training faster than GANs, have slow CPU inference. In this paper, we introduce WaveNeXt 2, a unified ConvNeXt-based framework compatible with both GAN and diffusion vocoders. Its core innovation is residual denoising and sub-modeling, where each sub-model progressively refines the waveform. Experimental results in the multi-speaker dataset demonstrate the effectiveness of our approach: (1) GAN-WaveNeXt 2 is much faster than HiFi-GAN and WaveFit, and (2) Diff-WaveNeXt 2 also delivers much faster inference and competitive synthesis quality compared with FastDiff with 4 steps. The Diff-WaveNeXt 2 is very training-efficient, training in only 32 hours, making it ideal for resource-constrained applications.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Audio, speech and music, one of 14 directions in Diffusion models. Builds on 23 of its 30 references that are in the map. First posted on arXiv 25 May 2026.

Builds on

23 of 30 references.

Closest ancestors · 7

Same direction, newest first

  1. 2022Hierarchical Diffusion Models for Singing Voice Neural Vocoder14 Oct
  2. 2022WaveFit: An Iterative and Non-autoregressive Neural Vocoder based on Fixed-Point Iteration3 Oct
  3. 2022FastDiff: A Fast Conditional Diffusion Model for High-Quality Speech Synthesis21 Apr
  4. 2022SpecGrad: Diffusion Probabilistic Model based Neural Vocoder with Adaptive Noise Spectral Shaping31 Mar
  5. 2022BDDM: Bilateral Denoising Diffusion Models for Fast and High-Quality Speech Synthesis25 Mar
  6. 2021PriorGrad: Improving Conditional Denoising Diffusion Models with Data-Dependent Adaptive Prior11 Jun
  7. 2020DiffWave: A Versatile Diffusion Model for Audio Synthesis21 Sep

Other references · 16

  1. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  2. 2022A ConvNet for the 2020scited by 3,579
  3. 2020HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesiscited by 1,154
  4. 2020WaveGrad: Estimating Gradients for Waveform Generationcited by 622
  5. 2022UTMOS: UTokyo-SaruLab System for VoiceMOS Challenge 2022cited by 390
  6. 2019MelGAN: Generative Adversarial Networks for Conditional Waveform Synthesiscited by 354
  7. 2022BigVGAN: A Universal Neural Vocoder with Large-Scale Trainingcited by 251
  8. 2021NISQA: A Deep CNN-Self-Attention Model for Multidimensional Speech Quality Prediction with Crowdsourced Datasetscited by 192
  9. 2023Vocos: Closing the gap between time-domain and Fourier-based neural vocoders for high-quality audio synthesiscited by 153
  10. 2023LibriTTS-R: A Restored Multi-Speaker Text-to-Speech Corpuscited by 101
  11. 2021UnivNet: A Neural Vocoder with Multi-Resolution Spectrogram Discriminators for High-Fidelity Waveform Generationcited by 74
  12. 2020Multi-band MelGAN: Faster Waveform Generation for High-Quality Text-to-Speechcited by 63
4 more
  1. 2022iSTFTNet: Fast and Lightweight Mel-Spectrogram Vocoder Incorporating Inverse Short-Time Fourier Transformcited by 32
  2. 2020Fixed Point Strategies in Data Sciencecited by 16
  3. 2024FreGrad: Lightweight and Fast Frequency-aware Diffusion Vocodercited by 11
  4. 2024SpecDiff-GAN: A Spectrally-Shaped Noise Diffusion GAN for Speech and Music Synthesiscited by 3

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2022InferGrad: Improving Diffusion Models for Vocoder by Considering Inference in Training
  2. 2023Voice Conversion with Denoising Diffusion Probabilistic GAN Models
  3. 2025On the Design of Diffusion-based Neural Speech Codecs
  4. 2023Adversarial Training of Denoising Diffusion Model Using Dual Discriminators for High-Fidelity Multi-Speaker TTS
  5. 2025BridgeVoC: Revitalizing Neural Vocoder from a Restoration Perspective
  6. 2022DiffGAN-TTS: High-Fidelity and Efficient Text-to-Speech with Denoising Diffusion GANs
  7. 2026Wave-Trainer-Fit: Neural Vocoder with Trainable Prior and Fixed-Point Iteration towards High-Quality Speech Generation from SSL features ↗
  8. 2022GAN You Hear Me? Reclaiming Unconditional Speech Synthesis from Diffusion Models