his is a Speech Emotion Recognition system that classifies emotions from speech samples using deep learning models. The project uses four datasets: CREMAD, RAVDESS, SAVEE, and TESS. The model achieves an accuracy of 96% by combining CNN, LSTM, and CLSTM architectures, along with data augmentation techniques and feature extraction methods.
machine-learning deep-learning cnn lstm human-computer-interaction convolutional-lstm clstm speech-emotion-recognition ravdess-dataset speech-emotion-classification tess-dataset savee-dataset crema-d-dataset
-
Updated
Nov 22, 2024 - Jupyter Notebook