1. 제 목 : Towards Peer-to-Peer Human-Robot Interaction

2. 연 사 : Prof. Lynne Parker (University of Tennessee Knoxville, USA)

3. 일 시 : 2011년 10월 21일 (금) 16:30~

4. 장 소 : 301동 1512-2호

5. Biosketch :
Dr. Lynne Parker is Professor and Associate Head in the Department of Electrical Engineering and Computer Science at the University of Tennessee, Knoxville, where she directs the research of the Distributed Intelligence Laboratory and is the Director of the Center for Intelligent Systems and Machine Learning.Additionally, she holds an appointment as Adjunct Distinguished Research and Development Staff Member at Oak Ridge National Laboratory (ORNL), where she worked as a full time researcher for several years.Dr. Parker received her Ph.D. degree from the Massachusetts Institute of Technology (MIT), performing her research in MIT's Artificial Intelligence Laboratory, with a minor in brain and cognitive science. Dr. Parker is recognized as a leading international expert in the area of multi-robot systems a field that she helped initiate with her dissertation work on this topic.Dr. Parker’s dissertation research (1994) on ALLIANCE, an architecture for multi-robot cooperation, was the first Ph.D. dissertation on the topic of multi-robot systems, and is considered a pioneering work in the field.Dr. Parker is an active leader in her field, and is a frequent invited speaker at international conferences, workshops, industries, and universities, having given over 100 invited lectures.
Dr. Parker has published over 125 articles in the areas of mobile robot cooperation, human-robot cooperation, robotic learning, intelligent agent architectures, and robot navigation.For this research, she was awarded the PECASE (U.S. Presidential Early Career Award for Scientists and Engineers) in 2000.She is the Editor-in-Chief of the IEEE Robotics and Automation Society Conference Editorial Board, and is an elected Administrative Committee (AdCom) Member of the IEEE Robotics and Automation Society.She was a senior Editor of IEEE Transactions on Robotics for several years, and is on the Editorial Board of IEEE Intelligent Systems and the Swarm Intelligence Journal.She is a Fellow of IEEE.

6. Abstract :
In peer-to-peer human-robot teaming, a primary objective is to create a style of cooperation between robot(s) and human(s) that is reminiscent of well-practiced human-only teams. In these human-only teams, the individuals have trained together, and understand intuitively how to interact with each other on the current task without the need for any explicit commands or conversations.In this interaction, the human performs tasks in a very natural manner, as he or she would when working with a human teammate; there is no need for PDAs, heads-up-displays, computers, or other types of graphical user interfaces (GUIs) to enable the human to communicate with the robot.This talk will introduce our recent work toward achieving this type of peer-to-peer human-robot interaction.Two primary thrusts of research will be introduced.The first addresses the problem of human activity recognition.In this research, we have proposed new 4-dimensional local spatio-temporal features that combine both intensity and depth information for the purpose of recognizing activities in typical indoor office environments.The second research thrust studies the problem of how to determine the intent of the human's actions by observing the effects of the human's actions.In this work, we use Conditional Random Fields to classify human actions, specifically in the context of a box pushing domain.We tie the work together by discussing how this research fits into a larger vision of peer-to-peer human-robot interaction.

7. 문 의 : 기계항공공학부 이동준 교수 (880-7114)