Speech Emotion Recognition
Machine Learning
Deep Learning
Speech Emotion Recognition
AI Application
Tech Stack
Python
TensorFlow
Convolutional Neural Networks (CNN)
Flask
REST APIs
Librosa
NumPy
Description
speech emotion recognition system built using deep learning techniques to classify human emotions from audio signals.
A Convolutional Neural Network (CNN) was trained on extracted speech features and achieved an accuracy of 86% on validation data. The project focused on effective audio preprocessing, model optimization, and inference performance.
The trained model was deployed using a Flask backend and exposed through RESTful APIs, enabling real-time emotion prediction and seamless integration with external applications.
- Built a CNN-based emotion recognition model for speech signals with 86% accuracy.
- Implemented audio preprocessing and feature extraction using Librosa.
- Deployed the trained model via a Flask backend with REST APIs.
- Enabled real-time emotion classification through a web-accessible service.
- Designed the system for easy integration into larger AI-driven platforms.