Project RoadMap

[codeperfectplus]

Voicem 03-12-2022

Create an Audio Authentication Library. Roadmap

Steps:-

1. Voice Preprocessing
2. Feature Extraction
3. Enrol/Auth
4. PyPI Package Deployment

Voice Possible Preprocessing

Should we extract the feature by ourselves or should try a deep learning approach?

1. Clear background and foreground voice.
2. Trim the part with no sound from the foreground
3. Extract feature after trimming voice
4. Create a feature stack with other possible features.

Deep Learning/Machine Learning for classification?

• Machine Learning Models
  • gaussian mixture
  • SVM
• Deep Learning Models
  • CNN Model