Project Summary
Human actions—such as drinking, reaching, or throwing—are composed of several smaller movement primitives. Understanding these basic sub-movements helps researchers analyze motor behavior, predict future actions, and teach robots to learn natural human motion.
In this project, you will develop an intelligent algorithm that automatically decomposes a complex human motion trajectory into its fundamental components. Using recorded 3D motion data, your task is to detect key turning points, segment the trajectory into meaningful motion units, and characterize each segment by direction and function.
What You Will Do
-
Collect or use provided 3D arm-motion datasets
-
Preprocess and analyze motion trajectories
-
Design algorithms to segment and classify movement primitives
-
Visualize and evaluate the decomposition results
Why This Project Matters
The ability to break down complex motions is essential for:
-
Human–robot interaction
-
Rehabilitation and sports analysis
-
Robot learning from human demonstrations
Your work will help build systems that better understand and predict human movement patterns.