Gender-Classifier-Using-Voice

audio source (10.4gb) = http://www.repository.voxforge1.org/downloads/SpeechCorpus/Trunk/Audio/Main/16kHz_16bit/

Total repository size = 33mb (approx)

SVM classifier model saved with 97% test accuracy to 'finalised_model.sav' and Neural Network classifier model saved to 'neural_network.h5' using this dataset. This has been created for distributing and testing on different languages,accents and on different people to compare different models.

Packages required - (pip install)

1. librosa
2. matplotlib
3. pyaudio
4. sklearn
5. pandas
6. scipy
7. numpy
8. wave
9. keras

You can now also intall and use pitch as library function. *pip install pitch* .

### Steps 1. Download the full "ML_final" folder only, at a specific path (preferred C:\\ ) 2. Open 'svm.py' or 'neural_network.py' and set the path variable to your downloaded folder (path="C:\\\\ML_final\\\\") 3. Run the 'svm.py' file and test your results for rbf kernal SVM model or try 'neural_network.py' for neural network model

Team

Atulya Kumar
Viren Baria
Bhargav Desai
Sanjeet Krishna
Parth Mehta

License and copyright

Licensed under the MIT license