CALL FOR POSTERS ===================
RSS 2016 Workshop on Robot-Environment Interaction for Perception and Manipulation: Interactive Perception Meets Reinforcement Learning and Optimal Control =================== Website: http://rss16ip-rl-oc.robotics.usc.edu Important Dates : Submission deadline: 29 April 2016 Notification of acceptance: 6 May 2016 Workshop date: 19 June 2016 =================== We invite interested researchers to submit abstracts to the RSS 2016 workshop on Robot-Environment Interaction for Perception and Manipulation. ABSTRACT Robots operating in real-world environments must tackle complex perception and manipulation tasks. Exploratory interaction with the environment is an efficient or even necessary way to obtain task-relevant information. Such information is contained in the way the sensor signals change in correlation with the robot’s actions. This knowledge is critical for understanding the state of the world (i.e, perception) and for learning how to change this state (i.e., manipulation). Interacting with the environment has been studied in several, mostly disjoint communities: dual and optimal control, reinforcement learning, and interactive perception. Interactive perception - leveraging a robot's actions to obtain and exploit information about complex environments - has usually relied on basic, highly-abstracted actions. In contrast, optimal control and reinforcement learning find optimal policies in complex manipulation and locomotion tasks but have often relied on simplifying assumptions about the robot’s perception. We believe that a tight coupling between manipulation and perception is vital to addressing manipulation tasks in partially observable environments. In particular, it allows gathering informative sensory signals through complex interactions with the environment and will enable accomplishing more perceptually-demanding manipulation tasks. OBJECTIVES The goal of this workshop is to bring together researchers from the interactive perception, optimal control, and reinforcement learning communities to discuss the aforementioned limitations and common problems. By combining these communities’ strengths, we hope to progress towards robotic systems that operate in partially observable environments and tackle tasks that are both perceptually demanding and require complex policies. We are looking forward to discussing future trends and open problems. CALL FOR ABSTRACTS Participants are invited to submit abstracts related to key challenges in robot-environment interaction, particularly focused on coupling perception and manipulation. Topics include but are not limited to: (i) Interactive Perception, (ii) Optimal Control, (iii) Reinforcement Learning. We invite submissions in the form of extended abstracts (up to 2 pages) following RSS formatting guidelines. The abstracts will be reviewed by the organizers. Accepted contributions will be featured in a poster session and will be included in the workshop proceedings, which will be available at the workshop webpage (http://rss16ip-rl-oc.robotics.usc.edu/). We encourage work-in-progress to be submitted and will take this into account in the review process. Submissions and questions should be directed to hausmanka...@gmail.com before 29th of April 2016. Please include "2016 RSS poster submission" in the subject of the email. Notifications of acceptance will be given by May 6th 2016. DATE AND LOCATION June 19th, 8:30-12:00 Rackam Auditorium at the University of Michigan Ann Arbor, Michigan WEBSITE AND SCHEDULE All information regarding the workshop will be published on the website http://rss16ip-rl-oc.robotics.usc.edu/. Soon, the workshop schedule and the list of invited speakers will be posted there as well. ORGANIZING COMMITTEE Karol Hausman, University of Southern California Herke van Hoof, TU Darmstadt Nikolay Atanasov, University of Pennsylvania Roberto Martin-Martin, TU Berlin Oliver Brock, TU Berlin
_______________________________________________ uai mailing list uai@ENGR.ORST.EDU https://secure.engr.oregonstate.edu/mailman/listinfo/uai
