Weight Balance Identification

We investigated a novel biometric that verifies people based on their unique weight balance and pressure distribution on the “Wii Balance Board” for four poses using a self-collected database of 80 subjects. Also, we implemented and studied a model based on PCA(Principal Components Analysis), LDA(Linear Discriminant Analysis) and DTW(Dynamic Time Warping). We validated it using Cross Validation at the best 10.5% of the Equal Error Rate. 

As a result, we successfully verified the hypothesis that ”Weight Balance” could be used as a new biometric. We won the Best Laboratory Intern Award at Yonsei University for this project.