Special Issue Focus

Autonomous robots are expected to perform a wide variety of everyday and specialised tasks in dynamic environments populated by humans and other artefacts. To perform human-centric collaborative tasks in such unstructured environments, robots will need to combine actions in intelligent ways to accomplish long-horizon, unseen tasks, while also communicating their intentions and capabilities to the humans with whom they share the environment. To enable this capability, robots must be endowed with knowledge of which actions they can perform, and an ability to reason about their consequences: both key elements of high-level cognitive behavior.

In contemporary robotics research, actions are interpreted in two main ways: first, as control policies responding to low-level sensor data; and second, as high-level symbolic actions. Action semantics can bridge these two levels, informing not just what to do but how to do it, and enabling effective human-robot collaboration in addition to autonomy. Recent advances in large-scale, general representation learning in computer cognitive vision, commonsense reasoning, and natural language processing indicates that the learned, general purpose action semantics for robotics are on the immediate horizon. Deep semantic representation and reasoning mediated visual perception and action provides a tool for capturing the essence and function of actions, thereby helping robots learn and generalize across task and motion planning domains. High-level learned semantic action representations will yield robots with greater capability and autonomy in a wide range of naturalistic human environments.

This special issue aims to collect the most prominent research results in the important and growing area of action semantics. This includes the state-of-the-art in generic action representation and reasoning at the intersection of Vision, AI, and Robotics communities, particularly looking for a common ground to combine different approaches for generalizable autonomy. The special issue will particularly highlight new methods allowing robots to learn generalized semantic models for different domains, as well as scalability and adaptation of the learned models to new scenarios/domains.

Key Topics of interest

• Task and Motion Planning

• Explainable and Interpretable Robot Decision-Making methods

• Active and Context-based Vision

• Cognitive Vision and Perception - Semantic Representations

• Commonsense reasoning about space and motion (e.g., for policy learning)

• Task-oriented and Perception-informed Language Grounding

• Task and Environment Semantics

• Robot Learning from Demonstration and Exploration

We particularly invite controbutions addressing:

— How can we learn scalable and general semantic representations? In recent years, there has been a substantial contribution in semantic policy and action representation in the fields of robotics, computer vision, and machine learning. In this respect, we would like to invite experts in academia and motivate them to comment on the recent advances in semantic reasoning by addressing the problem of linking continuous sensory experiences and symbolic constructions to couple perception and execution of actions. In particular, we want to explore how these can make robot learning more scalable and generalizable to new tasks and environments.

— How can semantic information be used to create Explainable AI? We would like to invite researchers from a broad range of areas including task and motion planning, language learning, general-purpose machine learning, and human-robot interaction. Much of action semantics is definitionally tied to how robots and humans communicate, and one fundamental feature of these approaches should be that they allow a broad variety of people to benefit from advances in robotics, and to work alongside robots outside of laboratory environments. Building more understandable action representations is important as a way of building robotic systems that benefit society.

If you would like to discuss the suitability of your work for the special issue, please send us an email using the contact details below.

APPLICABLE DATES IN 2022

• Paper submissions open (through Elsevier system): Dec 1 2021

• Final paper submission deadline: April 17, 2022

Reviews of submitted papers will commence as the papers are submitted. Earlier submissions may expect an overall quick turn-around time. As a worst case, we expect all accepted publications to be published in 2022.

Novelty Criteria for the SPAR-SI submissions

All submissions will be reviewed as regular RAS journal papers, and the journal criteria on novelty, originality apply.

We do welcome extended versions of published conference papers, or works resulting from the consolidation of several conference papers under the following criteria:

• The submitted papers must have at least 35% new impacting technical/scientific material in the submitted journal version.

• The submitted papers should have less than 50% verbatim similarity, e.g., as reported by a tool such as CrossRef.

• If applicable, authors should ensure to include a summary of key differences/novelties with respect to previously published work as part of the journal submission to the SPAR special issue; such previous works of the submitting authors should also be explicitly cited in the submitted article.

Submit your paper to the SPAR Special Issue

Please follow the next link to submit your manuscript to the SPAR Special Issue.

Note, that you need to choose the "Article Type" option VSI: SPAR 2022 from the site. The Elsevier portal will be open on the 1st of December to accept submissions to this Special Issue.

www.editorialmanager.com/robot/default1.aspx

Contact: Please direct all inquiries pertaining to the special issue to spar.workshop@gmail.com

Special Issue Editors

• Karinne Ramirez-Amaro (Chalmers, Sweden)

• Chris Paxton (NVIDIA, United States)

• Jesse Thomason (University of Southern California, United States)

• Maru Cabrera (University of Massachusetts Lowell, United States)

• Mehul Bhatt (Örebro University, Sweden)

Contact: Please direct all inquiries pertaining to the special issue to spar.workshop@gmail.com