# auspicious3000/autovc

AutoVC: Zero-Shot Voice Style Transfer with Only Autoencoder Loss

Repository: https://github.com/auspicious3000/autovc
Canonical: https://ross.abutalabs.com/products/autovc
Homepage: https://arxiv.org/abs/1905.05879
Language: Python
License: MIT
License Family: permissive
Topics: voice-conversion, speech-synthesis, generative-models, tacotron-pytorch, wavenet-vocoder, unsupervised-learning
Last push: 2024-10-23T00:53:39+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2668, "days_push": 680, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1100, forks 215 (observed 2026-08-28T04:03:35.419671+00:00)

## What it is
AUTOVC is a PyTorch implementation of a many-to-many non-parallel voice conversion framework that performs zero-shot voice style transfer using only an autoencoder with a carefully designed bottleneck. It includes pre-trained models, notebook-based conversion pipelines, and a WaveNet/HiFi-GAN vocoder for waveform synthesis.

## Use cases
- convert a voice recording to sound like a target speaker
- perform zero-shot voice conversion without target speaker training data
- do many-to-many voice conversion from non-parallel speech data
- retrain a voice conversion model on my own speech dataset
- synthesize waveforms from mel-spectrograms with a vocoder
- reproduce the ICML 2019 AutoVC paper results

## When to choose
- you need research-grade non-parallel or zero-shot voice conversion in PyTorch
- you want pre-trained models to convert voices without training from scratch
- you are reproducing or building on the AutoVC paper

## When to avoid
- you need real-time or production voice conversion with modern quality
- you want a maintained tool with active support
- you need text-to-speech rather than voice-to-voice conversion

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, audio-processing, speech-recognition
- domain: speech-processing, machine-learning, artificial-intelligence
- platform: python
- tags: voice-conversion, speech-synthesis, autoencoder, zero-shot, pytorch, wavenet-vocoder, style-transfer, audio

## Member repositories
- auspicious3000/autovc (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:35.419671+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:45:46.697794+00:00, confidence not recorded.
  - readme: https://github.com/auspicious3000/autovc (fetched 2026-08-28T04:03:35.419671+00:00, sha e6d1a2a6b429)
  - homepage: https://arxiv.org/abs/1905.05879 (fetched 2026-08-29T12:49:15.787393+00:00, sha 34e16ccbc0db)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T12:49:15.796889+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T12:49:15.801441+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T12:49:15.803380+00:00, sha a1f16f915a9a)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T12:49:15.799001+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.