Speech-emotion-recognition

Tools Required

Python 2.7 is used during development and following libraries are required to run the code provided.

Tensorflow
Numpy
Matplotlib
Librosa

Dataset

The IEMOCAP dataset used for model training. Another related dataset that can be used is Google's AudioSet.

The dataset is organized as indicated bellow: Fold1 - angry Fold2 - exc Fold3 - fea Fold4 - fru Fold5 - happy Fold6 - neu Fold7 - sad fold8 - sur