Speech Emotion Detection - Live Project
Developed and deployed an LSTM-based speech emotion recognition system using audio data. Enhanced the accuracy of emotion classification through advanced data preprocessing and MFCC feature extraction. Delivered a web application for emotion detection, making the tool accessible and user-friendly. • Implemented model training and evaluation for multiple emotions in speech data. • Conducted comprehensive data cleaning and feature engineering to improve performance. • Utilized Streamlit for deployment and user interaction enhancement. • Achieved an 82% accuracy rate, validating the quality of the labeled dataset.