The following code uses Hidden Markov Model for speech recognition.
- python=3.9
- numpy==1.18.2
- scikit-learn==0.22.2
- hmmlearn==0.2.4
- python-speech-features==0.6
Install all the dependencies and cd into the directory where this repo is and then run
python speech_recognizer.py --input-folder data/
The data folder contains audio datasets that are to be trained where the last file of each folder in data is used to test the accuracy of the model.