andybi7676/reborn-uasr

REBORN: Reinforcement-Learned Boundary Segmentation with Iterative Training for Unsupervised ASR

27
/ 100
Experimental

This project helps speech researchers and language model developers transcribe speech into phoneme sequences without needing extensive labeled audio data. It takes raw audio recordings, such as those from the LibriSpeech datasets in English or various other languages, and outputs the underlying phonemic structure of the speech. This is useful for building speech recognition systems or analyzing speech sounds.

No commits in the last 6 months.

Use this if you need to analyze or transcribe audio data in a language where labeled speech data for traditional ASR training is scarce or unavailable.

Not ideal if you already have a large, labeled dataset for your target language and can train a supervised ASR model.

speech-recognition phoneme-transcription language-modeling unsupervised-learning audio-analysis
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 5 / 25
Maturity 16 / 25
Community 6 / 25

How are scores calculated?

Stars

14

Forks

1

Language

Python

License

MIT

Last pushed

Dec 11, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/andybi7676/reborn-uasr"

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