Deep learning-based facial-gesture analysis system for motorized wheelchair control. Final BSc Computer Science Project, UPSIN (2018).
- Adrián Constante
- Brandon Vizcarra
- Erick Sámano
Deep Chair uses computer vision and deep learning to analyze facial-gesture posture in real-time to control wheelchair mottors and directions, providing feedback for movement..
- 📹 Real-time Analysis: Live facial-gesture detection from webcam
- 🧠 Deep Learning: CNN-based facial-gesture classification
- 🎯 Accuracy: 94% classification accuracy
Input (Webcam) → Preprocessing(OpenCV) → CNN Model(ResNet18) → Classification → Feedback
- PyTorch
- OpenCV
- NumPy/Pandas
- Matplotlib
MIT License