Autonomous Rubik’s Cube Solver
Fine-tuned the YOLOv8 model for cube face recognition, achieving 93% classification accuracy under varying lighting conditions. Engineered a robotic arm with 1 degree of freedom and integrated it with the recognition model, enabling the system to solve 46 out of 50 cubes autonomously. Developed a Unity-based GUI to simulate solving steps in real time, improving debugging efficiency and demonstration clarity. Published the work as a conference paper at IOE Graduate Conference (IOEGC), showcasing innovation in applied computer vision and robotics.



