TrafficVisionBD (Ongoing): AI-Powered Traffic Management for Bangladesh
Initiating a computer vision project to tackle traffic management challenges in Bangladesh, where road congestion results in over 3 million lost working hours daily.
Currently in the data collection phase, aiming to gather and annotate a dataset of 30,000+ traffic images from key urban areas to ensure model accuracy and reliability.
Futuregoals include developing an object detection model capable of identifying 20+ vehicle types and road elements (e.g., signs, barriers, risky zones), as well as performing time-series analysis to monitor traffic flow and predict high-risk situations for potential accident prevention.
Smart Identification Management System: (web based)
Developed a web-based smart identification system, utilizing HTML, CSS, PHP, and XAMPP to create an efficient and seamless ID management process.
Addressed the task of improving user management by designing features that streamline identification verification.
Deployed the system in real-time scenarios, reducing processing times and improving overall user satisfaction with the system's functionality