This codebase is for the project 'Speech Emotion Recognition using CNN', and is prepared by Shubhankar Borse.

Take a look at the report file attached here for more information about this project.



 Dear User,

 Once you go through the report for the project and you wish to simulate the results, please follow these steps.


 1) Download the RAVDESS and SAVEE audio databases and extract al .wav files to the folder RawData. Add this to the folder containing this ReadMe 
 file. This is the required dataset.

 
 

2) Install the libraries: Keras, Librosa, Tensorflow, pandas, sklearn, glob, scipy on your python3 environment.



 3)There are two ways to validate results : Either run the feature extraction and CNN training .py file 'Train_The_Model.py' or 

 you can also use the pretrained network model on the test feature set which is present in this folder by directly running 'Test_The_Model.py' 


 Regards,

 Shubhankar Borse