Create Your First Project
Start adding your projects to your portfolio. Click on "Manage Projects" to get started
Blink Detection AI for Driver Safety
Location
Human Systems Lab, University of Windsor
Date
JUNE 2024 - AUGUST 2024
Project type
SELF DRIVING, DRIVER DISTRACTION
Assessment
Take over time, cognitive workload
Link
At Human System Labs, I developed a convolutional neural network (CNN) using NVIDIA TinyCUDA to detect driver blinks in real-time, enhancing road safety. This project processed live video feeds to monitor blink frequency and duration, indicators of driver fatigue or inattention. By integrating machine vision with AI, I created a robust system to alert drivers to potential safety risks, advancing human-machine interaction for safer driving experiences.
Key Contributions:
-Built a CNN-based blink detection model using NVIDIA TinyCUDA for real-time video processing.
-Analyzed blink patterns to assess driver fatigue and attention levels.
-Developed a system to provide real-time safety alerts, improving driver awareness.
-Ensured system reliability through rigorous testing and data validation.
