Contrôle de la marche des robots humanoïdes complexes par apprentissage par imitation

The project is part of a Master’s degree Thesis titled “Imitation Learning for Gait Similarity Optimization in Bipedal Locomotion Robots”, which aims to study the application of artificial intelligence techniques to allow humanoid robots to learn how to walk in a way as similar as possible as a human does. Specifically, the project to be developed in UQTR addresses the application of imitation learning algorithms and optimization methods, in conjunction with motion capture data of human locomotion. These motion data describe gait patterns that are compared to those of the robot, so that through a set of specific parameters, the difference is minimized. This optimization process is performed by means of multiple tests, both in a simulation environment and in a real robot (available at the host university), so that the differences between human dynamics and the humanoid robot can be studied and eventually managed.

Faculty Supervisor:

Alben Cárdenas

Student:

Partner:

Universidad Nacional de Colombia

Discipline:

Engineering

Sector:

Technology; Artificial Intelligence

University:

Université du Québec à Trois-Rivières

Program: