ADHD Diagnosis via Deep Learning Video Analysis
Bachelor's thesis- Python
- PyTorch
- TensorFlow
- OpenCV
- CNN
- Eye Gaze Tracking
Developed an AI-driven system for detecting atypical behaviors to assist in ADHD diagnosis using video processing and Deep Learning.
- Created a custom video dataset of subjects performing concentration-specific tasks.
- Implemented a Convolutional Neural Network (CNN) to detect and analyze sudden, atypical head movements.
- Developed eye gaze tracking methods to evaluate subject attention levels during tasks like reading.
- Performance benchmarked against existing diagnostic methods to evaluate system efficiency and costs.
Achieved 9.90 out of 10 on my thesis.