Telegram-Zalo/zac2022-lyric-alignment

Solution for Zalo AI Challenge 2022 - Lyrics Alignment

43
/ 100
Emerging

Combines source separation (Demucs) with fine-tuned Wav2Vec2 acoustic modeling and dynamic programming-based forced alignment to map lyrics to precise audio timestamps. The pipeline includes data augmentation through random segment merging, custom vocabulary adaptation for Vietnamese, and character-level CTC training followed by viterbi-style alignment to extract word-level temporal boundaries. Designed for the Zalo AI Challenge 2022 with reproducible training scripts and checkpoint downloads.

No commits in the last 6 months.

Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 8 / 25
Maturity 16 / 25
Community 19 / 25

How are scores calculated?

Stars

68

Forks

18

Language

Python

License

Apache-2.0

Last pushed

Dec 05, 2022

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/Telegram-Zalo/zac2022-lyric-alignment"

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