primepake/wav2lip_288x288

Wav2Lip version 288 and pipeline to train

53
/ 100
Established

Implements multi-resolution lip-sync generation (288x288 to 512x512) using SAM-UNet architecture with advanced training techniques including Wasserstein loss, gradient penalty, and PReLU/LeakyReLU activations. Two-stage training pipeline: first trains a SyncNet for audio-visual synchronization, then fine-tunes the Wav2Lip generator with attention-based upsampling. Builds on the original Wav2Lip framework with architectural improvements for higher-quality facial dubbing synthesis.

642 stars. No commits in the last 6 months.

Stale 6m No Package No Dependents
Maintenance 2 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 25 / 25

How are scores calculated?

Stars

642

Forks

162

Language

Python

License

MIT

Last pushed

Aug 13, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/primepake/wav2lip_288x288"

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