alecokas/BiLatticeRNN-Confidence

Confidence Estimation for Black Box Automatic Speech Recognition Systems Using Lattice Recurrent Neural Networks https://arxiv.org/abs/1910.11933 or https://ieeexplore.ieee.org/document/9053264

36
/ 100
Emerging

This tool helps improve how accurately you can tell if an Automatic Speech Recognition (ASR) system made a mistake on a specific word. It takes an ASR system's output, whether a single best guess or a network of competing transcription options, and provides better confidence scores for each word. An ASR system developer or researcher would use this to refine their system's reliability.

No commits in the last 6 months.

Use this if you need more precise and reliable confidence scores for individual words produced by your black-box ASR system.

Not ideal if you are looking for an off-the-shelf ASR system or a solution that doesn't require technical understanding of neural networks and data processing.

speech-recognition ASR-quality-assurance natural-language-processing speech-tech-development
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 5 / 25
Maturity 16 / 25
Community 15 / 25

How are scores calculated?

Stars

14

Forks

4

Language

Python

License

MIT

Last pushed

Apr 16, 2020

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/alecokas/BiLatticeRNN-Confidence"

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