Statistical Modeling of Human Motion

Julian Faraway

Department of Statistics, University of Michigan

Friday, April 18

101 Yost Hall
Talk: 4:00 -- 5:00 p.m.
Refreshments: 3:30 -- 4:00 p.m. in 327 Yost

Motion capture technology can be used to collect data on human motion. Using a database of thousands motions, consisting of the trajectories of the joints, for a variety of standing individuals of various anthropometries performing lifting tasks to a range of targets, we show how to build a regression model for predicting these motions. The statistical interest lies in the novelty of the data where the regression responses are multivariate functions describing the orientation of the body segments or curves in 3D. These elements must be coherently combined to predict the motion of an individual with any anthropometry to precisely place an object on a target. An application to Ergonomics will be demonstrated.


If you have any questions, contact Ramani S. Pilla or Sharon Dingess