important dates

April 15, 2014
Paper submission

April 30, 2014
Notification of acceptance

June 1, 2014
Workshop at ICRA 2014

Invited Talks

Dr. Lucas Paletta

Computational Human Factors Analysis and Preferred Looks

Human factors have been central to psychology research since the 1950ies, as physical or cognitive properties of an individual or social behavior specific to humans that may influence the functioning of technological systems. This talk firstly highlights on how computational models of human factors have already entered the scientific debate in human-robot interaction, by giving examples from joint attention processes, measurements of collaboration, user requirements from mental processes, and contextual settings from workload and stress indicators. This will lead further to a discussion on preferred looks within the frame of the author's recent work on 3D gaze recovery, real-time human factors monitoring as well as performance evaluation in mobile interaction design. The debate is then contrasted from a discussion on how interaction requirements of social target groups (elderly, migrants, ASC users, etc.) introduce profundity into human factors research, in particular on computational modeling of attention. The talk concludes with a tentative outlook on future directions in computational human factors analysis, within the context of human-robot interaction.

Prof. Olivier Stasse

Vision based motion generation for humanoid robots

In this presentation, some results on planning vision based whole-body motion for humanoid robot will be presented. First a high level behavior, namely treasure hunting, is introduced to state the problems related to motion generation for humanoid robots. The goal is to plan tasks for ensuring locally robustness through sensor loop, and at the same time maintaining a feasible path, avoiding collisions and using at best the robot motion capabilities. This is achieved first by building high-level motion controllers which are able simultaneously to generate walking gaits and find footsteps. Some recent extensions which also minimize a vision based criteria will be presented. In a second time, planning algorithms to find sequences of motions ensuring visibility will be introduced. Finally a new way to qualify the feasibility of whole body motion controller and to take into account the environment is presented.