Tai-Chi-Grading-System

To further explore the field of AI, I am now developing a grading system for T’ai Chi Ch’uan (a traditional Chinese martial art) using Kinect, OpenNI and OpenGL as my final-year research project. . In this research project, I have used Kinect to capture and record the T’ai Chi actions performed by experts and users, reduced image dimensions using the Principal Component Analysis (PCA), and saved experts’ actions and characteristic information in database. The follow-up step I am now working on is to use Naïve-Bayes-Nearest-Neighbor (NBNN) classier to match users’ actions with experts’, and to covert T’ai Chi rules from text to geometry for grading. Although challenging, this AI research project is promising in that it helps establish a standard, objective grading system to avoid human factors (e.g. referees’ fatigue), and can also be used as an application for the training and dissemination of T’ai Chi Ch’uan as a cultural heritage among the public.