In search for the motion primitives for a communicative human body language

This PhD project is part of MoPrim (In Search for the Motion Primitives for a Communicative Human Body Language). The aim of this project is to investigate ways which allow automatic finding of a set of human body language motion primitives. The motion primitives must be of such quality that they can be used for both automatic recognition and synthesis of normal human body language. This project will focus on the human upper body, and has therefore limited the motion primitives to only this area. Preliminary tests and prototype developments have already begun, and will in 2005 lead to a final choice of method. This project aims at building systems for both automatic construction of a set of motion primitives and synthesis of human body language based on the primitives, in 2006. (Lars Reng, Erik Granum, Thomas B Moeslund)