https://github.com/mjhydri/singing-vocal-beat-tracking

This repo contains the source code of the first deep learning-base singing voice beat tracking system. It leverages WavLM and DistilHuBERT pre-trained speech models to create vocal embeddings and trains linear multi-head self-attention layers on top of them to extract vocal beat activations. Then, it uses HMM decoder to infer signing beats and tempo.
https://github.com/mjhydri/singing-vocal-beat-tracking

beat-tracking hubert linear-transformer music music-information-retrieval self-supervised singing-voice wavlm

Last synced: 12 months ago
JSON representation

This repo contains the source code of the first deep learning-base singing voice beat tracking system. It leverages WavLM and DistilHuBERT pre-trained speech models to create vocal embeddings and trains linear multi-head self-attention layers on top of them to extract vocal beat activations. Then, it uses HMM decoder to infer signing beats and tempo.

• Host: GitHub
• URL: https://github.com/mjhydri/singing-vocal-beat-tracking
• Owner: mjhydri
• License: mit
• Created: 2022-04-06T17:58:38.000Z (about 4 years ago)
• Default Branch: main
• Last Pushed: 2022-09-04T19:41:59.000Z (almost 4 years ago)
• Last Synced: 2025-07-05T06:11:38.329Z (12 months ago)
• Topics: beat-tracking, hubert, linear-transformer, music, music-information-retrieval, self-supervised, singing-voice, wavlm
• Language: Python
• Homepage:
• Size: 122 KB
• Stars: 33
• Watchers: 2
• Forks: 4
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# Singing Beat Tracking With Self-supervised Front-end and Linear Transformers

The source code is available in this repo!

The readme file will be added soon!