rishikksh20/Fre-GAN-pytorch

Fre-GAN: Adversarial Frequency-consistent Audio Synthesis

45
/ 100
Emerging

This tool helps researchers and audio engineers synthesize high-quality audio from various input parameters, ensuring the generated sound maintains consistent frequency characteristics. It takes in acoustic features, often derived from speech or other audio, and produces realistic, natural-sounding audio waveforms. This is ideal for those working on advanced audio generation tasks.

111 stars. No commits in the last 6 months.

Use this if you need to generate very high-fidelity, natural-sounding audio that precisely matches the frequency characteristics of your input acoustic features.

Not ideal if you're looking for a complete, end-to-end voice cloning or text-to-speech solution, as this focuses specifically on the audio synthesis (vocoder) component.

audio-synthesis speech-technology sound-engineering voice-generation acoustic-modeling
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 9 / 25
Maturity 16 / 25
Community 20 / 25

How are scores calculated?

Stars

111

Forks

31

Language

Python

License

MIT

Last pushed

Aug 26, 2021

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/rishikksh20/Fre-GAN-pytorch"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.