Human sensorimotor learning for robot skill synthesis
Prof. Erhan Oztop
ATR, Japan
Time: 13th September

Abstract: Primates, in particular, humans are very adept at learning to use tools. In this talk, I will introduce a paradigm that utilizes this sensorimotor learning capacity to obtain robot behaviors, which otherwise would require manual programming by experts. The idea is to consider the target robotic platform as a tool that can be controlled by a human. Provided with an intuitive interface for controlling the robot, the human learns to perform a given task using the robot. This is akin to the stage where a beginner is learning to drive a car. After sufficient learning, the skilled control of the robot by the human provides learning data points that are used for constructing an autonomous controller so that the robot can perform the task without human guidance. I will demonstrate the feasibility of this framework by presenting several examples including a manipulation skill obtained for a robotic hand, and statically stable reaching skill obtained for a small humanoid robot.

From an engineering point of view, this paradigm relies on techniques from teleoperation and machine learning, and has the same goals with robotic imitation and robot learning by demonstration. The key difference is that the proposed paradigm includes the human in the control loop and employs the human brain as the adaptive controller to accomplish a given task. Once the control proficiency has been attained by the human, then obtaining an autonomous controller boils down to reverse engineering the control policy established by the human brain.

As time permits, I will present some ideas on the neural correlates of human-in-the-loop robot control and show how the interfaces built for robot skill synthesis can also be used in the reverse direction for probing motor control mechanisms employed by the central nervous system.

Speaker bio: Erhan Oztop earned his Ph.D. at the University of Southern California in 2002. In the same year, he joined the Computational Neuroscience Laboratories at Advanced Telecommunications Research Institute International (ATR) in Japan where he worked as a researcher until 2004. Immediately after, he joined the JST ICORP Computational Project as a researcher, and later became a group leader (2004-2008). Currently, he is a senior researcher of ATR Cognitive Mechanisms Laboratories where he is leading the Communication and Cognitive Cybernetics group. He holds a visiting associate professor position at Osaka University and is also affiliated with NICT, Biological ICT group.

His research interests include computational and cognitive neuroscience, human-robot interaction and computational modeling and analysis of intelligent behavior.
 

Vision for Acting and Interacting Robots
Prof. Danica Kragic
School of Computer Science and Communication
Royal Institute of Technology (KTH)
Time: 14th September

Abstract: The ability to autonomously acquire new knowledge through interaction with the environment is one of the major research goals in the field of robotics. The knowledge can be acquired only if suitable perception-action capabilities are present.  In other words, a robotic system has to be able to detect, attend to and manipulate objects in the environment. In the first part of the talk, we present the results of our longterm work in the area of vision based sensing and control. The work on finding, attending, recognizing and manipulating objects in domestic environments is discussed. More precisely, we present a stereo based active vision system framework where aspects of foveated attention are put into focus and demonstrate how the system can be utilized for object grasping.

The second part of the talk presents work on the visual analysis of human manipulation actions which are of interest for e.g. human-robot interaction applications where a robot learns how to perform a task by watching a human. A method for classifying manipulation actions in the context of the objects manipulated, and classifying objects in the context of the actions used to manipulate them is presented.  The action-object correlation over time is then modeled using conditional random fields. Experimental comparison shows improvement in classification rate when the action-object correlation is taken into account, compared to separate classification of manipulation actions and manipulated objects.

Speaker bio: Danica Kragic is a Professor at the School of Computer Science and Communication at KTH in Stockholm. She received MSc in Mechanical Engineering from the Technical University of Rijeka, Croatia in 1995 and PhD in Computer Science from KTH in 2001. She has been a visiting researcher at Columbia University, Johns Hopkins University and INRIA Rennes. She is the Director of the Centre for Autonomous Systems at KTH and the coordinator for the EC integrated project GRASP.

Danica received the 2007 IEEE Robotics and  Automation Society Early Academic Career Award. She is charing the IEEE RAS Technical Committee on Computer and Robot Vision and from 2009 serves as an IEEE RAS AdCom member. Her research is in the area of computer vision, object grasping and manipulation and human-robot interaction. In the area of computer vision she is interested in developmnet of active vision systems allowing robots to interact with humans and each others in realistic environments. Her recent work explores different learning methods for formalizing the models for integrated representation of objects and actions that can be applied on them. This work has demonstrated how robots can achieve scene understanding through active exploration and how full body tracking of humans can be made more efficient.
 

Mediated perception and action in Human Robotic Embodiment
Prof. Massimo Bergamasco
PERCRO
Scuola Superiore S.Anna (SSSA), Italy
Time: 15th September

Abstract: Research on Robotics and Artificial Intelligence presently interprets the interaction between human and robotic subjects through a set of mediated controls who engage almost all human senses. The underlying communication requires the establishment of a set of interaction metaphors in order to allow a natural type of control. The most common of these metaphors is the kinematic copy of human gestures.

During the last ten years incredible progresses in cognitive science, psychology, neuroscience, physiology and neurophysiology have been achieved. The knowledge about the relationship between the "perception and sensori-motion mechanisms" and the brain has never been so advanced. One of the major outcomes is in the fact that it is possible to correlate brain neuronal activities to corresponding perception and motion, and, at the same time, it is possible to record these activities through the use of innovative devices. The combined adoption of these results does allow dissolving the physical constraints that act as a boundary during the interaction and control of robots and virtual environment entities.

A new generation of brain and body (computer) interfaces (BBCI) will allow direct transfer of user intention and the realization of remote sensing. In such a way it will be possible to achieve the interactive communication with a robot without the requirement to physically perform any action. Virtual and robotic bodies can be controlled and moved even in absence of a correspondent motion of the controlling subject. The robots is moved only by thoughts while the robot perception is transferred directly to the humans through worn interfaces (Head mounted displays, skin and body stimulators,…)

This lecture addresses the path from teleoperation and virtual environment interaction toward new methods to recreate the illusion of surrogating our own bodies in different entities (being robotic or virtual) and investigates how the relevant perception-action loops will be affected.

Speaker bio: Massimo Bergamasco is Full Professor of Applied Mechanics at Scuola Superiore Sant’Anna, where he teaches Mechanics of Robots and Virtual Environments. His research activity deals with the study and development of haptic interfaces for the control of the interaction between humans and Virtual Environments. He has been the Scientific coordinator of several National projects and EU projects. The scientific activity of Massimo Bergamasco includes more than two hundreds scientific papers published on journals and/or international conferences proceedings. Massimo Bergamasco is member of the Editorial Board of the several international Journals and Conferences of robotics and computer graphics, He has been a member of the Steering Committee of the Interaction Design Institute, Ivrea Italy. Massimo Bergamasco has been the Coordinator of the ENACTIVE Integrated Project and is presently a cofounder of the European Institute on Enactive Systems (EIES). Massimo Bergamasco is presently the coordinator of the EU IP SKILLS aiming at exploring the use of multimodal interfaces to transfer human skills across subjects