Skip to content

Latest commit

 

History

History
2 lines (2 loc) · 496 Bytes

README.md

File metadata and controls

2 lines (2 loc) · 496 Bytes

emodbSER

This project aims to realize Speech Emotion Recognition task on speech data of Berlin Database of Emotional Speech using deep convolutional neural network technique. With the help of autoencoder technique to pretrain the network, compared to the Convolutional Neural Network without pretraining our structure could improve the recognition accuracy from 20% to 60%. The Neural Networks are implemented based on Keras framework.