2
1
Looking for experience!
This Flutter-based project is an AI-powered Voice Emotion Detection system that classifies human emotions from voice inputs using deep learning. Users can upload or record audio through an integrated Gradio web interface. The app extracts Mel-Frequency Cepstral Coefficients (MFCCs) using the librosa library to capture key features of the voice signal. These features are processed by a pre-trained Multilayer Perceptron (MLP) model built with TensorFlow/Keras, which identifies emotions like happy, sad, angry, and neutral. The predicted emotion is displayed in real-time within the Flutter UI. This app can be used in virtual assistants, customer support, mental health tracking, and interactive systems.
1 May 2025