Worked for the first time with audio data where I learned how the features is extracted from audio data using MFCC — Mel-Frequency Cepstral Coefficients, and then using these features I trained a sequential model with 6 hidden layers.
Dataset can be downloaded from : https://drive.google.com/file/d/1g64EswaS5PtwIg-Y0ZmWwvSK1DgYvUuc/view?usp=sharing Download the .ipynb notebook to implement the code only run the cells which:
- installs the libraries
- imports the libraries
- the get_MFCC function
- last 5 cells after the markdown comment "Testing the Model with your own Voice"
Save the notebook in the same folder where male_clips folder and female_clips folder is there. Save the results folder (Which has the model.h5 file) in the same folder where the notebook is stored (i.e. where male_clips folder and female_clips folder is there.)