Our latest research focuses on automating construction using self-interlocking building blocks and robotic arms. This project explores high-precision tracking, impedance control, and simulation to achieve robust and efficient assembly. You can read more about it in our blog post: Teaching Robots to Build
AR Tower Defense game we developed as a project in class. We use OpenCV for visualization and pose calculation. To detect the gestures, we use Google MediaPipe for hand segmentation and then evaluate the detected keypoints for different gestures.