mlsanityData to 19 Oct 2026

PhaseGAN: High-Fidelity Vocoder via Decoupled Amplitude and GAN-Driven Phase Reconstruction

arXiv 2609.12918 ↗First posted 11 Sep 2026cs.SDcited by 0

On the map: Speech and audio language models → Vocoders and speech synthesis

Abstract (arXiv)

A vocoder is a pivotal component of modern text-to-speech (TTS) systems. Despite the significant progress of neural network-based vocoders, accurate phase reconstruction remains the main challenge limiting both audio quality and modeling efficiency. We introduce PhaseGAN, a lightweight vocoder that addresses this limitation through a "mel rightarrow Amplitude rightarrow Phase" reconstruction pipeline. By reconstructing amplitude and phase spectra via distinct methodologies, the proposed PhaseGAN outperforms state-of-the-art baselines while utilizing fewer model parameters and reduced computational requirements. The compact version generates high-fidelity audio with approximately 500K parameters and 1 GMAC computational load, making it highly suitable for real-time applications on edge devices. In addition, our approach exhibits exceptional musical audio synthesis capabilities despite no training on musical data, illustrating unprecedented cross-domain generalization. See https://github.com/phasegan/phasegan-audio-demo for demos of our work.

Key facts. No citing papers in the map yet (data through 19 Oct 2026). Placed in Vocoders and speech synthesis, one of 11 directions in Speech and audio language models. Builds on 27 of its 43 references that are in the map. First posted on arXiv 11 Sep 2026.

Builds on

27 of 43 references.

Closest ancestors · 3

Same direction, newest first

  1. 2024FreeCodec: A disentangled neural speech codec with fewer tokens2 Dec
  2. 2024FreeV: Free Lunch For Vocoders Through Pseudo Inversed Mel Filter12 Jun
  3. 2022iSTFTNet: Fast and Lightweight Mel-Spectrogram Vocoder Incorporating Inverse Short-Time Fourier Transform4 Mar

Other references · 24

  1. 2016SGDR: Stochastic Gradient Descent with Warm Restartscited by 5,126
  2. 2017Decoupled Weight Decay Regularizationcited by 2,114
  3. 2016WaveNet: A Generative Model for Raw Audiocited by 1,880
  4. 2020HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesiscited by 1,462
  5. 2020DiffWave: A Versatile Diffusion Model for Audio Synthesiscited by 1,394
  6. 2017Natural TTS Synthesis by Conditioning WaveNet on Mel Spectrogram Predictionscited by 1,023
  7. 2023ConvNeXt V2: Co-designing and Scaling ConvNets with Masked Autoencoderscited by 737
  8. 2022UTMOS: UTokyo-SaruLab System for VoiceMOS Challenge 2022cited by 552
  9. 2019MelGAN: Generative Adversarial Networks for Conditional Waveform Synthesiscited by 436
  10. 2018WaveGlow: A Flow-based Generative Network for Speech Synthesiscited by 365
  11. 2022BigVGAN: A Universal Neural Vocoder with Large-Scale Trainingcited by 315
  12. 2018Efficient Neural Audio Synthesiscited by 280
12 more
  1. 2017Parallel WaveNet: Fast High-Fidelity Speech Synthesiscited by 240
  2. 2023Vocos: Closing the gap between time-domain and Fourier-based neural vocoders for high-quality audio synthesiscited by 218
  3. 2020DDSP: Differentiable Digital Signal Processingcited by 190
  4. 2017Stabilizing Training of Generative Adversarial Networks through Regularizationcited by 119
  5. 2022Opencpop: A High-Quality Open Source Chinese Popular Song Corpus for Singing Voice Synthesiscited by 65
  6. 2025The GAN is dead; long live the GAN! A Modern GAN Baselinecited by 61
  7. 2022Avocodo: Generative Adversarial Network for Artifact-free Vocodercited by 24
  8. 2023APNet: An All-Frame-Level Neural Vocoder Incorporating Direct Prediction of Amplitude and Phase Spectracited by 10
  9. 2023APNet2: High-quality and High-efficiency Neural Vocoder with Direct Prediction of Amplitude and Phase Spectracited by 8
  10. 2020Phase reconstruction based on recurrent phase unwrapping with deep neural networkscited by 7
  11. 2024Low-Latency Neural Speech Phase Prediction based on Parallel Estimation Architecture and Anti-Wrapping Losses for Speech Generation Taskscited by 3
  12. 2025Neural Vocoders as Speech Enhancerscited by 3

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2024FA-GAN: Artifacts-free and Phase-aware High-fidelity GAN-based Vocoder
  2. 2022PhaseAug: A Differentiable Augmentation for Speech Synthesis to Simulate One-to-Many Mapping
  3. 2025High-Fidelity Music Vocoder using Neural Audio Codecs
  4. 2026Prosody-Guided Harmonic Attention for Phase-Coherent Neural Vocoding in the Complex Spectrum
  5. 2023HiFTNet: A Fast High-Quality Neural Vocoder with Harmonic-plus-Noise Filter and Inverse Short Time Fourier Transform
  6. 2024ScoreDec: A Phase-preserving High-Fidelity Audio Codec with A Generalized Score-based Diffusion Post-filter
  7. 2022HiFi-WaveGAN: Generative Adversarial Network with Auxiliary Spectrogram-Phase Loss for High-Fidelity Singing Voice Generation
  8. 2024VNet: A GAN-based Multi-Tier Discriminator Network for Speech Synthesis Vocoders