Intro

My name is Zaid Khan, a junior computer science student at Kennesaw State University. I have been interested in computers my whole life and took an interest in software engineering when I started university. I compete in robotics competitions such as VEXU and NHRL. Feel free to check out what I have worked on.

My goal is to leverage advanced computational techniques to solve real-world problems and contribute to the field of software. My journey in computer science started with a curiosity for how machines think and evolved into a deep dive into other various technologies. I am always eager to learn more and apply my expertise to solve challenging problems and push the boundaries of what I know. Outside of my academic and professional life, I enjoy exploring new technologies, reading, and working with cars.

Projects

Over the course of my studies at KSU, I've had the opportunity to work on various projects that showcase my skills and dedication to artificial intelligence, machine learning, and software development. These projects demonstrate my ability to innovate, solve complex problems, and deliver impactful solutions. Notable highlights include:

  • Portfolio Website: Created an interactive web app using Heroku, Flask, HTML, Tailwind, and Python, allowing users to simulate results on their datasets and images, showcasing practical applications of my work.
  • Perceptron: Developed a single-layer neural network using Python and Numpy to predict outcomes ('bright' or 'dim') from binary input values as part of a supervised learning task.
  • Autonomous Robotics: Engineered autonomous navigation algorithms in C++ for VEXU robots, enabling precise field-oriented positioning and contributing to a top-3 worldwide placement.
  • Image Processing Tools: Implemented pixel-level operations using Python, Pillow, Numpy, and OpenCV for tasks such as downscaling, quantization, transformations, and histogram equalization.
  • Live Traffic Sign Detection: Designed a traffic sign recognition system using a ResNet-152 CNN model trained on the GTSRB dataset, achieving high accuracy across 43 traffic sign classes and integrating real-time video processing via OpenCV.
  • Diabetic Retinopathy Classification: Built a fine-tuned ResNet-50 model in PyTorch for classifying diabetic retinopathy in retinal images, achieving over 98% accuracy with optimized data augmentation techniques.
  • E-Commerce Website: Developed a database-driven e-commerce platform using C#, Entity Framework, and SQL Server, featuring user account management, shopping cart functionality, and transactional records.