A research group made of Assistant Professor Hideki Kadone of the University of Tsukuba Center for Cybernetics Research, in collaboration with Associate Professor Gentiane Venture of the Tokyo University of Agriculture and Technology, and Professor Emeritus Alain Berthoz of the Collège de France, have extracted and mapped individual emotional characteristics from human gait data (the way a person walks) into a model they used to show that it is mathematically possible to recognize emotions. The research team created the model based on data obtained from actual people walking, and studied the parameters that influence the recognition of emotions. They then built a database from the results they obtained, and used emotional recognition algorithms they created from that database to verify the possibility of emotional recognition. These algorithms suggest the possibility that human emotions can be quantitatively predicted.

Original Paper

Gentiane Venture, Hideki Kadone, Tianxiang Zhang, Julie Grèzes, Alain Berthoz, Halim Hicheur, Recognizing Emotions Conveyed by Human Gait, International Journal of Social Robotic 6(4), doi:10.1007/s12369-014-0243-1(25 Jun 2014)

Figure : A model created from motion capture data of an actor's emotional expressions