Séance Séminaire

Séminaire de Probabilités et Statistique

Bio-inspired tools for analyzing dynamics and motor signature of patient with schizophrenia in a context of interaction with a social assistive robot (Séminaire KIM)

The development of robotics has recently made it possible to integrate robots as therapeutic tools. Socially assistive Robotics in particular is very promising for developing therapy, especially for people suffering from certain psychiatric disorders. Our context here is the exploitation of robots as tools to encourage people with schizophrenia to exercise. To develop and improve the behavior of these robots, it is important that they can adapt to the patient. The objective of this post-doctorate is thus the development of methods to detect the motor characteristics of a patient (motor signature, inter and intra personal coordination) from video. This is done with the long-term perspective of exploiting these tools for the improvement of socially assistive robots. The originality of our approach is that, rather than using classic image processing tools, we exploit a bio-inspired neural model which is based on the exploitation of rhythmic training phenomena. This model permits to analyze the dynamics of a subject in a video. Such an approach allows the development of motor control capable of handling multiple frequencies and degrees of freedom with, for application, a robot capable of producing complex and coordinated movements with a patient.

WEBINAIRE ouvert à toutes et tous : https://umontpellier-fr.zoom.us/j/85813807839

ATTENTION: Séminaire KIM - débute à 14h.