Speed Limit Sign Detector

Designed and implemented a computer vision system using Support Vector Machines (SVM) and Convolutional Neural Networks (CNN) in Python to detect speed limit signs from real-time video feeds. The project included data pre-processing, model training with a custom dataset, and integration with ROS2 for vehicle control, significantly enhancing the autonomous vehicle’s navigation systems.

YOLOv11 Object Detection System

Developed a real-time object detection and tracking system using the YOLOv11 pre-trained model. The system captures live frames from a webcam, detects objects, optimizes performance using Region of Interest (ROI), and tracks object trajectories to predict future positions. The goal was not only to demonstrate effective object detection but also to understand the impact of performance optimizations through detailed logging and visualizations.

Bayesian Personalized Ranking model

Developed a sophisticated machine learning model using Bayesian inference techniques to analyze and predict user preferences with high accuracy. Employed Python's Jupyter Notebook for iterative testing and tuning, utilizing libraries such as Pandas and NumPy for data manipulation and Matplotlib for visualization. The model successfully improved the personalization of movie recommendations by learning from implicit user feedback and achieved significant improvement in AUC performance over multiple epochs.

Ruby on Rails Web-Application

A comprehensive web application using Ruby on Rails, React.JS, and JavaScript, implementing robust authentication mechanisms, event management systems, and service logging features. Developed a scalable backend infrastructure with Ruby on Rails, integrating RESTful APIs and database models for efficient data management. Utilized React.JS to create dynamic and responsive user interfaces, ensuring optimal user experience and accessibility.