DNR-ROB 2021: Declarative and Neurosymbolic Representations in Robot Learning and Control Virtual Event July 15-16, 2021 |
Conference website | https://dnr-rob.github.io |
Submission link | https://easychair.org/conferences/?conf=dnrrob2021 |
Submission deadline | June 27, 2021 |
RSS 2021 Workshop on Declarative and Neurosymbolic Representations in Robot Learning and Control
A Full-day Virtual Event
July 12-16, 2021 (To be determined)
https://dnr-rob.github.io/
Important dates
Paper Submission Deadline: June 20th, 2021 June 27th, 2021, Anywhere on Earth (UTC-12)
Acceptance Notification: June 30th, 2021, Anywhere on Earth (UTC-12)
Workshop date: one day between July 12th - 16th, 2021 (TBD)
About:
This is a joint workshop with two themes that share common interests and motivations.
Declarative knowledge in learning and control of robot behaviors: For the purposes of explainability, abstraction, efficiency, or robustness, declarative knowledge is being incorporated into the decision-making process. We aim to explore novel ways to leverage complementary features of the different forms of decision making in order to inform the future research of deliberative systems relying on declarative knowledge that can be learned from, and shared with, humans.
Neurosymbolic robotics for learning symbolic representations from sub-symbolic representations: Neurosymbolic AI has emerged to integrate successful ideas in deep learning and classical symbolic reasoning in a single framework. Such a framework will have the desirable perception abilities of deep networks for bottom-up computation whilst allowing the system to make symbolic reasoning for top-down computation.
In this sense, both themes are essentially interested in possible ways of the unification of multiple methodologies to create robotics systems that are scalable, explainable, efficient, and robust. We intend to bring together:
- Robotics researchers from the once distant fields of knowledge representation and reasoning, symbolic and motion planning, reinforcement learning, and more generally machine learning for behavior recognition and synthesis
- Robotics researchers who are interested in incorporating human knowledge in declarative forms into robot behaviors
- Robotics researchers who work on symbolic AI approaches and would like to incorporate recent advances in deep learning
- Robotics researchers who work on deep learning and would like to exploit the reasoning and explainability related capabilities of symbolic manipulation systems
- Researchers in other AI fields (such as vision and natural language processing) who work in the intersection of declarative, neural, and symbolic approaches
- Cognitive and developmental roboticists who work on symbol grounding and emphasize symbol emergence
Invited Speakers
- Masataro Asai (IBM Research)
- Anthony Cohn (University of Leeds)
- Katerina Fragkiadaki (CMU)
- Nick Hawes (University of Oxford)
- Leslie Pack Kaelbling (MIT)
- George Konidaris (Brown University)
- Ben Kuipers (University of Michigan)
- Luís C. Lamb (Federal Uni. of Rio Grande do Sul)
- Cynthia Matuszek (UMBC)
- Sheila McIlraith (University of Toronto)
Submission Guidelines
We will accept regular papers (up to 8 pages) and extended abstracts (up to 4 pages) in standard RSS format, excluding unlimited pages for references. The review process will be in a single-blind fashion. The plan is to allocate 10 minutes for presenting each regular paper, and 2 minutes for each extended abstract paper.
Paper Submission Deadline: June 27th, 2021
Paper Acceptance Notification: June 30th, 2021
Organizing Committee
- Alper Ahmetoglu (contact person), Bogazici U, ahmetoglu.alper@gmail.com
- Shiqi Zhang (contact person), SUNY Binghamton, zhangs@binghamton.edu
- Roderic Grupen, UMass Amherst
- Yuqian Jiang, UT Austin
- Matteo Leonetti, U of Leeds
- Tiffany Liu, UMass Amherst
- Erhan Oztop, Ozyegin U, Osaka U
- Justus Piater, U of Innsbruck
- Benjamin Rosman, U of Witwatersrand
- Tadahiro Taniguchi, Ritsumeikan U
- Emre Ugur, Bogazici U
Contact
All questions about submissions should be emailed to zhangs@binghamton.edu and alper.ahmetoglu@boun.edu.tr