2024 DroGone An autonomous drone-catching drone. Python ROS Computer Vision Deep Learning +2 DroGone 2024 An autonomous drone-catching drone. A cutting-edge autonomous system that uses computer vision and deep learning to detect, track, and capture rogue drones. Features real-time flight path prediction and precision interception capabilities. Technologies: Python ROS Computer Vision Deep Learning Object Detection Flight Control Visit Project → 2023 Hole Detection in Robotics Computer vision system for automated hole detection in manufacturing robotics at Microsoft. Python PyTorch Computer Vision YOLO +2 Hole Detection in Robotics 2023 Computer vision system for automated hole detection in manufacturing robotics at Microsoft. Sophisticated machine vision system leveraging deep learning and computer vision techniques for automated hole detection, surface defect inspection, and quality control in manufacturing environments. The project employs convolutional neural networks (CNNs), advanced image processing algorithms, and real-time analysis to identify, classify, and measure holes, surface defects, and manufacturing anomalies with sub-millimeter precision. Designed for industrial quality control applications in automotive, aerospace, and precision manufacturing industries where component integrity is critical. Technologies: Python PyTorch Computer Vision YOLO OpenCV Manufacturing Visit Project → 2023 AADefDINO Exploring Adversarial Attacks and Defenses in Vision Transformers trained with DINO. PyTorch Vision Transformers Adversarial ML DINO +2 AADefDINO 2023 Exploring Adversarial Attacks and Defenses in Vision Transformers trained with DINO. Research project investigating adversarial robustness in self-supervised Vision Transformers. Implemented novel attack methods and defense strategies for DINO-trained models, contributing to adversarial ML security. Technologies: PyTorch Vision Transformers Adversarial ML DINO Self-Supervised Learning Research Visit Project →