Real-Time Face Emotion Recognition System
A Deep Learning application using custom CNN architectures for real-time emotion analysis and automated attendance tracking.
Developed during a software engineering internship, this project focuses on Computer Vision and Deep Learning to analyze human emotions in real-time. The system integrates facial recognition with emotion classification to create an automated attendance and monitoring solution.
Technical Approach
The core of the system relies on Convolutional Neural Networks (CNNs) to process video feeds:
- Custom CNN Architecture: Designed and trained a custom CNN model to classify emotions (Happy, Sad, Angry, Neutral) with high precision.
- Performance: Achieved 87% accuracy in real-time testing environments.
- Face Verification: Utilized DeepFace and OpenCV for robust face detection and alignment before emotion analysis.
System Features
- Real-Time Processing: Capable of processing video streams with low latency for immediate feedback.
- Automated Attendance: Replaced manual tracking methods by identifying registered users and logging their entry times automatically.
- Interactive GUI: Built a user-friendly interface using Tkinter to visualize detection results and manage user data.
Tech Stack
- Frameworks: TensorFlow, Keras, DeepFace
- Computer Vision: OpenCV
- Language: Python
- Interface: Tkinter
The system demonstrates the practical application of Deep Learning in security and administrative automation.