Emotional speech is trained using the following datasets:

TESS: https://www.kaggle.com/datasets/ejlok1/toronto-emotional-speech-set-tess

DAIC-WOZ: https://dcapswoz.ict.usc.edu/daic-woz-database-download/

For facial expression classification, DeepFace Library is used

The Research paper published for this project can be found at: https://ijcrt.org/papers/IJCRT22A6776.pdf