Methods › Computer Vision › Generative Adversarial Networks › HiFi-GAN
HiFi-GAN
Introduced by Jungil Kong et al. in HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis
archive 2025-07-28 Description, source and code snippet are the archive's method entry.
HiFi-GAN is a generative adversarial network for speech synthesis. HiFi-GAN consists of one generator and two discriminators: multi-scale and multi-period discriminators. The generator and discriminators are trained adversarially, along with two additional losses for improving training stability and model performance.
The generator is a fully convolutional neural network. It uses a mel-spectrogram as input and upsamples it through transposed convolutions until the length of the output sequence matches the temporal resolution of raw waveforms. Every transposed convolution is followed by a multi-receptive field fusion (MRF) module.
For the discriminator, a multi-period discriminator (MPD) is used consisting of several sub-discriminators each handling a portion of periodic signals of input audio. Additionally, to capture consecutive patterns and long-term dependencies, the multi-scale discriminator (MSD) proposed in MelGAN is used, which consecutively evaluates audio samples at different levels.
In Syntology repository known, no verified implementation (declared_code; repositories: 5 listed of 11 in the graph, strongest link: official). This answer is about arXiv:2010.05646, the paper the archive credits with this method.
Code Syntology holds under other names, as the lookup lists them (code_nearby); not a complete list, and none of them is this method.
- IT-GAN · shares_name_tokens (shares name tokens with this name; a different method) · level V3 · retrieve with
get_reference_implementation(name='IT-GAN') - GT-GAN · shares_name_tokens (shares name tokens with this name; a different method) · level V2 · retrieve with
get_reference_implementation(name='GT-GAN') - Diffusion-GAN · shares_name_tokens (shares name tokens with this name; a different method) · level V3 · retrieve with
get_reference_implementation(name='Diffusion-GAN') - Self-Diagnosing GAN · shares_name_tokens (shares name tokens with this name; a different method) · level V2 · retrieve with
get_reference_implementation(name='Self-Diagnosing GAN') - Reciprocal CF GAN (RCF-GAN) · shares_name_tokens (shares name tokens with this name; a different method) · level V3 · retrieve with
get_reference_implementation(name='Reciprocal CF GAN (RCF-GAN)') - variational GAN training framework · shares_name_tokens (shares name tokens with this name; a different method) · level V3 · retrieve with
get_reference_implementation(name='variational GAN training framework')
Origin paper of this answer, arXiv:2010.05646 (its page on this site · its page on Syntology): Syntology's run record for its harvested code, 17 ran (of which 11 constructed an object rather than computing a result; 14 with no instrument failure: 0 honoured, 0 violated, 14 with no contract checked; 3 where Syntology's instrument failed) · 8 unverified (of 25 harvested samples) · 3 pointer-only (licence). The instrument figure counts failures of Syntology's instrument, not of the code. Ran means executed on a synthesized input, not that the code is correct.
Syntology's have check for this name (free, it answers coverage only), read 2026-10-05; ask it again live. A level is the served verification level of that one answer; nothing here is summed across methods.
What each level means, quoted from Syntology's server card: V0 “imports”; V1 “executes against a synthesized fixture and returns a value of the recorded type”; V2 “two implementations GENERATED from the paper's signature and spec agree numerically (1e-6) on fixture inputs” “neither of the two ever read the authors' own code, so V2 measures how unambiguously the paper specifies the method, not fidelity to the authors' implementation (measured 2026-09-11, commit 75eaa77: cross_checked_by is a second LLM on 2,628 of 2,831 served samples)”; V3 “paper-derived property tests pass on both of those same two generations.” The card adds: “No rung on this ladder compares the served code against the authors' implementation, so treat every level as evidence about runnability and internal agreement, not about matching the paper's results.” A level quoted in two parts is one definition the card writes as two clauses joined by a double hyphen, each clause quoted on its own (how these levels are quoted).
Papers archive 2025-07-28
archive papers tagged: 33 · with a code link: 18 · where Syntology ran a sample: 6 (6 with a run with no instrument failure, 0 where every run was a failure of Syntology's instrument) Syntology
Show: all tagged papersonly where code ran (6 of 33 tagged: 6 with a run with no instrument failure, 0 where every run was a failure of Syntology's instrument)
Page 1 of 1: papers 1 to 33 of 33, newest first by the archive's date (ties by slug), in archive order.
Repository counts are the archive's code-links table. A Syntology line states what Syntology ran from that paper's harvested code, as “N ran (of which C constructed an object rather than computing a result; K with no instrument failure: H honoured, V violated, P with no contract checked; I where Syntology's instrument failed) · U unverified”; the instrument figure counts failures of Syntology's instrument, not of the code. It is per sample and not a correctness claim. When the archive marks a repository official for the paper, the line starts with that repository's state (the archive's flag, not a verdict on who wrote the code; “community repositories only” when every sample that ran came from a community repository, “official: no sample here; runs from other or unrecorded repositories” when some came from a repository the paper names or has in its text, or from none recorded); hover it for the repositories the samples that ran came from.
-
RingFormer: A Neural Vocoder with Ring Attention and Convolution-Augmented Transformer 2 Jan 2025 · 1 repository · arXiv:2501.01182
-
A Unified Model For Voice and Accent Conversion In Speech and Singing using Self-Supervised Learning and Feature Extraction 11 Dec 2024 · 0 repositories · arXiv:2412.08312
-
TSELM: Target Speaker Extraction using Discrete Tokens and Language Models 12 Sep 2024 · 1 repository · arXiv:2409.07841
-
DSP-informed bandwidth extension using locally-conditioned excitation and linear time-varying filter subnetworks 22 Jul 2024 · 0 repositories · arXiv:2407.15624
-
StreamSpeech: Simultaneous Speech-to-Speech Translation with Multi-task Learning 5 Jun 2024 · 1 repository · arXiv:2406.03049Syntology official (archive's flag): 4 ran · 4 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 2 where Syntology's instrument failed) · 0 unverified (of 4 harvested samples)
-
CoVoMix: Advancing Zero-Shot Speech Generation for Human-like Multi-talker Conversations 10 Apr 2024 · 1 repository · arXiv:2404.06690
-
SpecDiff-GAN: A Spectrally-Shaped Noise Diffusion GAN for Speech and Music Synthesis 30 Jan 2024 · 0 repositories · arXiv:2402.01753
-
UNIT-DSR: Dysarthric Speech Reconstruction System Using Speech Unit Normalization 26 Jan 2024 · 0 repositories · arXiv:2401.14664
-
Scaling NVIDIA's Multi-speaker Multi-lingual TTS Systems with Zero-Shot TTS to Indic Languages 24 Jan 2024 · 0 repositories · arXiv:2401.13851
-
SELM: Speech Enhancement Using Discrete Tokens and Language Models 15 Dec 2023 · 0 repositories · arXiv:2312.09747
-
APNet2: High-quality and High-efficiency Neural Vocoder with Direct Prediction of Amplitude and Phase Spectra 20 Nov 2023 · 1 repository · arXiv:2311.11545
-
Collaborative Watermarking for Adversarial Speech Synthesis 26 Sep 2023 · 0 repositories · arXiv:2309.15224
-
HiFTNet: A Fast High-Quality Neural Vocoder with Harmonic-plus-Noise Filter and Inverse Short Time Fourier Transform 18 Sep 2023 · 1 repository · arXiv:2309.09493
-
Rep2wav: Noise Robust text-to-speech Using self-supervised representations 28 Aug 2023 · 0 repositories · arXiv:2308.14553
-
MusicLDM: Enhancing Novelty in Text-to-Music Generation Using Beat-Synchronous Mixup Strategies 3 Aug 2023 · 1 repository · arXiv:2308.01546Syntology official (archive's flag): 10 ran · 10 ran (of which 0 constructed an object rather than computing a result; 6 with no instrument failure: 2 honoured, 3 violated, 1 with no contract checked; 4 where Syntology's instrument failed) · 4 unverified (of 14 harvested samples) · 14 pointer-only (licence)
-
Speaker-independent neural formant synthesis 2 Jun 2023 · 0 repositories · arXiv:2306.01957
-
Source-Filter-Based Generative Adversarial Neural Vocoder for High Fidelity Speech Synthesis 26 Apr 2023 · 1 repository · arXiv:2304.13270
-
Wave-U-Net Discriminator: Fast and Lightweight Discriminator for Generative Adversarial Network-Based Speech Synthesis 24 Mar 2023 · 0 repositories · arXiv:2303.13909
-
Self-Supervised Representations for Singing Voice Conversion 21 Mar 2023 · 0 repositories · arXiv:2303.12197
-
Fast and small footprint Hybrid HMM-HiFiGAN based system for speech synthesis in Indian languages 13 Feb 2023 · 0 repositories · arXiv:2302.06227
-
MnTTS2: An Open-Source Multi-Speaker Mongolian Text-to-Speech Synthesis Dataset 11 Dec 2022 · 1 repository · arXiv:2301.00657
-
Hiding speaker's sex in speech using zero-evidence speaker representation in an analysis/synthesis pipeline 29 Nov 2022 · 1 repository · arXiv:2211.16065
-
Towards Building Text-To-Speech Systems for the Next Billion Users 17 Nov 2022 · 2 repositories · arXiv:2211.09536Syntology community repositories only · 1 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified (of 1 harvested sample)
-
Source-Filter HiFi-GAN: Fast and Pitch Controllable High-Fidelity Neural Vocoder 27 Oct 2022 · 0 repositories · arXiv:2210.15533
-
MnTTS: An Open-Source Mongolian Text-to-Speech Synthesis Dataset and Accompanied Baseline 22 Sep 2022 · 1 repository · arXiv:2209.10848
-
Mandarin Singing Voice Synthesis with Denoising Diffusion Probabilistic Wasserstein GAN 21 Sep 2022 · 0 repositories · arXiv:2209.10446
-
JETS: Jointly Training FastSpeech2 and HiFi-GAN for End to End Text to Speech 31 Mar 2022 · 2 repositories · arXiv:2203.16852Syntology official (archive's flag): 2 ran · 2 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified (of 2 harvested samples)
-
Disentangleing Content and Fine-grained Prosody Information via Hybrid ASR Bottleneck Features for Voice Conversion 24 Mar 2022 · 0 repositories · arXiv:2203.12813
-
ECAPA-TDNN for Multi-speaker Text-to-speech Synthesis 20 Mar 2022 · 1 repository · arXiv:2203.10473
-
iSTFTNet: Fast and Lightweight Mel-Spectrogram Vocoder Incorporating Inverse Short-Time Fourier Transform 4 Mar 2022 · 2 repositories · arXiv:2203.02395Syntology 11 ran (of which 0 constructed an object rather than computing a result; 5 with no instrument failure: 1 honoured, 0 violated, 4 with no contract checked; 6 where Syntology's instrument failed) · 0 unverified (of 11 harvested samples) · 6 pointer-only (licence)
-
YourTTS: Towards Zero-Shot Multi-Speaker TTS and Zero-Shot Voice Conversion for everyone 4 Dec 2021 · 3 repositories · arXiv:2112.02418
-
Text-Free Prosody-Aware Generative Spoken Language Modeling 7 Sep 2021 · 1 repository · arXiv:2109.03264
-
HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis 12 Oct 2020 · 11 repositories · arXiv:2010.05646Syntology official (archive's flag): 5 ran · 17 ran (of which 11 constructed an object rather than computing a result; 14 with no instrument failure: 0 honoured, 0 violated, 14 with no contract checked; 3 where Syntology's instrument failed) · 8 unverified (of 25 harvested samples) · 3 pointer-only (licence)
Tasks archive 2025-07-28
20 shown of 61 tasks the archive attaches to papers tagged with this method, by distinct papers. A task without a page in the catalog is plain text.
| Task | Papers |
|---|---|
| Speech Synthesis | 16 |
| Text to Speech | 13 |
| text-to-speech | 13 |
| Text-To-Speech Synthesis | 7 |
| Voice Conversion | 6 |
| Generative Adversarial Network | 5 |
| Audio Generation | 3 |
| Decoder | 3 |
| Automatic Speech Recognition (ASR) | 2 |
| CPU | 2 |
| Self-Supervised Learning | 2 |
| Speaker Verification | 2 |
| Speech Enhancement | 2 |
| Speech Recognition | 2 |
| Voice Cloning | 2 |
| speech-recognition | 2 |
| Automatic Speech Recognition | 1 |
| Bandwidth Extension | 1 |
| Beat Tracking | 1 |
| Data Augmentation | 1 |
Usage over time archive 2025-07-28
Components: the archive holds no method-to-method composition, so PwC's Components table cannot be rebuilt; the Papers list carries no Results column for the same reason (the archive does not join its leaderboard rows to method tags).
Categories archive 2025-07-28
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections