seungwonpark/melgan
MelGAN vocoder (compatible with NVIDIA/tacotron2)
This project helps convert text-based phonetic representations (mel-spectrograms) into natural-sounding speech. You input a mel-spectrogram, which is like a visual fingerprint of sound, and it generates the corresponding raw audio. This is ideal for researchers or developers working on text-to-speech systems, enabling them to quickly turn their speech models' outputs into audible sound.
650 stars. No commits in the last 6 months.
Use this if you need to transform mel-spectrograms generated by a text-to-speech model (especially one compatible with NVIDIA's Tacotron 2) into high-quality, synthetic speech.
Not ideal if you're looking for a complete text-to-speech system that takes text as input, or if you need to process audio formats other than mel-spectrograms.
Stars
650
Forks
114
Language
Python
License
BSD-3-Clause
Category
Last pushed
Oct 03, 2020
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/seungwonpark/melgan"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
kan-bayashi/ParallelWaveGAN
Unofficial Parallel WaveGAN (+ MelGAN & Multi-band MelGAN & HiFi-GAN & StyleMelGAN) with Pytorch
fatchord/WaveRNN
WaveRNN Vocoder + TTS
shangeth/wavencoder
WavEncoder is a Python library for encoding audio signals, transforms for audio augmentation,...
rishikksh20/iSTFTNet-pytorch
iSTFTNet : Fast and Lightweight Mel-spectrogram Vocoder Incorporating Inverse Short-time Fourier...
lucasnewman/best-rq-pytorch
Implementation of BEST-RQ - a model for self-supervised learning of speech signals using a...