The current methods in the literature for representing human motions are based on modeling the movements at a single level of abstraction, either at a low-level (i.e., trajectory level) of abstraction or at the high-level (i.e., symbolic level) of abstraction. The proposed project will exploit the latest advances in recurrent neural networks for modeling human motions at multiple hierarchical levels of abstraction. The ultimate aim is to allow patients to perform rehabilitation exercises at home using a sensory system for capturing the motions, where an algorithm will retrieve the trajectories of patient’s exercises, will perform a data analysis by comparing the performed motions to a model of desired motions, and will send the analysis results to the patient’s physician with recommendations for improvement.
Patel, Colleagues Find Pesticide May Contribute to Global Obesity EpidemicSeptember 13, 2021
U of I Molecular Modeler Jagdish Patel worked with Canadian scientists to screen several common food additives or contaminants. They discovered that the commonly sprayed organophosphate insecticide chlorpyrifos puts the break on […]
What’s your COVID-19 exposure risk in a gathering?November 24, 2020
Thank you to reporter Kyle Pfannenstiel for highlighting some of U of I’s COVID-19 modeling efforts, as originally published in the Post Register. If you’ll be at the dinner table […]
It Takes a Village (and a Research University)November 20, 2020
This article was written by Alexiss Turner, Marketing and Communications Manager from the College of Engineering, for the recently published “Here We Have Idaho” magazine. IMCI and many of our […]