label_fixer.py - Prepare labels to the correct format
sound_segmenter.py - Segment audio files using Audacity's simple label format
mfcc.py - MFCCs extraction from audio files
dtw.py - Simple naive implementation of the DTW algorithm for speech recognition
GMM-HMM.ipynb - Simple implementation of a GMMHMM model for speech recognition
To run:
- Clone the repository
- Install requirements -
pip install -r requirements.txt - Put audio and label data into
data/ - (Optional) Fix labels if they are in the extended format
- Segment the audio data (
sound_segmenter.py) - Run the algorithms
Demo:
