neosapience/mlp-singer

Official implementation of MLP Singer: Towards Rapid Parallel Korean Singing Voice Synthesis (IEEE MLSP 2021)

46
/ 100
Emerging

Implements a non-autoregressive, attention-free MLP-based architecture inspired by MLP-Mixer for parallel mel-spectrogram generation, achieving 200× real-time factor on CPU and 3400× on GPU. Requires input triplets of MIDI transcriptions, lyrics text, and monophonic audio; integrates HiFi-GAN as the neural vocoder backend for waveform generation, with support for alternative vocoders like WaveGlow.

118 stars. No commits in the last 6 months.

Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 20 / 25

How are scores calculated?

Stars

118

Forks

26

Language

Python

License

MIT

Last pushed

Feb 24, 2022

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/neosapience/mlp-singer"

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