Here I implement and study voice recognition methods. The idea is to implementation a variety of speaker recognition methods with explanations and results discussion.
Checklist of things done:
- Classic GMM gender-recognition approach
- Classic GMM voice-recognition approach
- GMM-UBM gender-recognition approach using MAP-adaptation
- GMM-UBM voice-recognition approach using MAP-adaptation
- Supervector based GMM-UBM voice-recognition approach
- GMM-UBM/i-vector approach
Currently datasets used are from Abhijeet Kumar's article. Addition to this dataset is now in development.