Graduate Projects

Some of the class projects that I’ve done before 2016 are listed on this page. For current research, please visit my PhD Research Page.

Experimental Game Technologies

This 8000 level course was a complete joy with Prof. Stephen Guy. With its detailed explanations of game techniques like One-Button Games, Crowd Simulations, advanced AI techniques, and visual and sound effects, it quickly became my favorite class of all times.

• For the One-Button Game project, I designed and coded a small boat game called Sea Rescue One-Button Game. The objective was to rescue as many people as possible to the shore line, while not crashing into rocks and other obstacles. Fuel and repairs had to be taken depending on the mileage and accidents. It also had sound effects (SFX).

• For the second game with “no-touching” criteria, I re-engineered my Head-Pose Estimation (HPE) code into a bubble bursting game. Player had to burst bubbles coming towards them by moving their head. It also had simple SFX.

• For my final project, I used Raspberry Pi 3.0 to create an Augmented Reality (AR) Tic-Tac-Toe game with the help of ArUCo marker detection library. It used a USB camera input and worked in real-time.

Virtual Reality

Virtual Reality is an ongoing research for me, but for this time period, I got to experiment with different edge-detection and Non-Photo Realistic (NPR) techniques like Sobel, Canny, Cartoon Shaders, Video Abstraction, Laplacian Pyramid, Difference of Gaussian (DoG), eXtended DoG (XDOG), and Flow-based DoG (F-DoG) both on CPU and GPU. I also modified Kyprianidis’ open-source X-DoG CUDA code to work in stereo for our custom video-see-through system.

Computer Vision

Computer Vision course with Prof. Volkan Isler was a serious but an exciting and fun class. Many computer vision techniques were explained and explored in details including edge-detections, bluring, blobs, corners and features detection, homography, image warping, transformations, SIFT, perspective-n-points, optical flow, image segmentation, snake active contours, 3D rectification of stereo pairs, and panaromas. For the final project with a little help, I chose the challenging Head-Pose Estimation (HPE) topic with a single calibrated webcam. It was magical to see how combining all these technologies works in a practical application.


Data Visualization class with Prof. Dan Keefe was an exciting experience among other CSci courses. It started like an art class, with crayons and sketches, and slowly dived into visualizing flows and scientific data. We learned many visualization and data filtering techniques like linking and brushing, zooming and scaling, navigations, and more. For the final project, I visualized the microbiome population in mice’s guts treated with antibiotics in D3JS. The data was given to me by Prof. Dan Knights, and it was available in each 1 millimeter section of the dissected gut. To make it a little more fun, I chose the name Gus’s Gut for this project.