|Schedule||Speakers||Call for Posters||Description|
|08:40-09:20||Benjamin Kuipers: Actions and Affordances: Among the Foundations for Commonsense Knowledge.|
|09:20-10:00||Hedvig Kjellström: Using Context for Better Object Functionality and Affordance Modeling.|
|11:10-11:50||Gorden Cheng: Enabling robots to observe, segment and recognize human activities from observations using semantic representations.|
|11:50-12:30||Florentin Wörgötter: Bootstrapping the Semantics of Tools: Affordance analysis of real world objects on a per-part basis.|
|14:10-14:50||Yiannis Demiris: Hierarchical action representations for assistive robotics|
|16:00–16:15||Functional Model of Everyday Objects for Handover Tasks; Natsuki Yamanobe et. al.|
|16:15–16:30||Extraction of Spatial Object Relations For Understanding the Semantics of Manipulation Actions; Fatemeh Ziaeetabar et. al.|
|16:30–16:45||A Representation for General Pouring Behavior; Akihiko Yamaguchi et. al.|
|16:45–17:00||See What I Mean—Probabilistic Optimization of Robot Pointing Gestures; Khurram Gulzar et. al.|
|17:10–18:00||Panel discussion, Poster session and Close|
Observation and execution are dual problems for autonomous robots. A sensory-motor bridge from observation to execution of an action is essential for autonomous robots to learn from human demonstrations and generate dynamic policies to fulfill demonstrated goals. This requires not only imitation of human movements at the level of continuous signals, but moreover, semantic action representation to express the desired goals even in different scene contexts. The more expressive and structural the semantic representation, the greater the capabilities and reliability of the robot.
We strongly believe that IROS 2015 is an ideal venue to shed light on recent successes and next steps in semantic action and policy representations for autonomous robots. From this workshop, we expect participating researchers to identify and address important challenges, techniques, and benchmarks.
The aim of this one-day workshop is two-fold. First, we will highlight recent developments in manipulation action semantic representations and semantic policy generation. We will also compare the state-of-the-art approaches for generic action and policy representations in both computer vision and robotics, looking for a common ground to combine assumedly disparate approaches for autonomous capability and reliability. A key goal is to reconcile and integrate various bottom-up and top-down approaches for semantic action perception.
Second, we will motivate our expert speakers to comment on (1) how to link symbolic representations with sensory experience while being fully grounded at the signal level to couple perception and execution of actions, and (2) how to open new avenues to build robots with greater capability and autonomy. Given these insights, we want to discuss important next steps and open problems in semantic action perception and policy learning.
This workshop is intended for roboticists working in the areas of perception, controls, planning, and learning, and who are interested in improving the reliability and autonomy of robots. We hope to bring together outstanding researchers and graduate students to discuss current trends, problems, and opportunities in semantic action (policy) representations, encouraging communication and common practices among scientists in this field.
We encourage 1-2 pages extended abstract of relevant work that has been previously published, or is to be presented at the main conference. The accepted abstracts will be posted on the workshop website in a compiled yearbook and will not appear in the official IEEE proceedings.
All papers must be written in English and submitted in PDF format. Each extended abstract has a maximum length of 2 pages.
Submissions must follow the IROS'2015 formatting instructions.
Reviewing is single blind, and will be carried out by the workshop chairs.
Please submit your abstract to the workshop official email address firstname.lastname@example.org before August 15, 2015.
Accepted abstract is expected to give a poster presentation at the workshop.
Outstanding abstracts are expected to give a 10-15 min talk at the workshop.
We encourage submissions that address the following topics:
Signal to symbol transition (Symbol grounding).
Semantics of manipulation actions.
Semantic policy representation.
Learning from demonstration.
Bottom-up and top-down perception.
Spatiotemporal event encoding.
Machine learning techniques for semantic representations.
Task, geometric, and dynamic level plans and policies.
Please submit a 1-2 page abstract in the IROS 2015 format to the organizers by August 15. Accepted posters will be notified by September 07.
This workshop is supported by the European projects Xperience, Poeticon++ and TimeStorm.