The focus of this Symposium is on the connection between abstract- and physical-level representations of objects in autonomous robotic systems. We call "anchoring" the process of creating, and maintaining in time, this connection. Anchoring can thus be seen as a special case of symbol grounding where the symbols denote physical objects.

Anchoring must necessarily occur in any physically embedded system that comprises a symbolic reasoning component. A typical example is the problem of connecting, inside an autonomous robot, the symbol used by a symbolic planner to refer to a physical object to the data in a perceptual system that pertains to the same object. This connection must be dynamic, since the same symbol must be connected to new percepts when the same object is re-acquired. For instance, a robot may be asked to identify and track a specific person in a crowd using visual data and given a linguistic description.

Anchoring must also occur in a multiple robot system whenever the robots exchange information via symbolic representations. We talk in this case of "grounded communication". A typical example is the problem of establishing the correspondence between the symbols used by two different robots embedded in the same physical environment to refer to the same physical object. In RoboCup, for instance, knowledge about the objects in the domain may need to be exchanged between robots. Grounded communication is also needed for efficient human-robot cooperation.

The main preoccupation of this Symposium is a practical one. Although all existing robotics systems that comprise a symbolic reasoning component implicitly incorporate a solution to the anchoring problem, this solution is typical hidden in the code, and it is developed on a system by system basis on a restricted domain. The ambition of this symposium is to create an interdisciplinary community that will develop a general theory of anchoring. In particular, we are interested in unveiling the computational aspects of anchoring, including the functionalities and representations needed to perform it. We believe that having such a theory will greatly advance our ability to build intelligent embedded systems, and to transfer techniques and results between different systems.

The Symposium includes invited talks, presentations of technical papers and position papers, and panel discussions. These pre-symposium proceedings collect the abstracts of the invited talks, plus the full text of the technical papers and position papers. The papers are available in either PDF or PostScript format. PostScript files are compressed using gzip. File size is indicated on the left.

Higher-order behavior-based systems,

Ian D. Hoswill

Northwestern University, Illinois, USA

Conceptual spaces: bridging symbolic, conceptual and connectionist,

Peter Gärderfors

Lund University, Sweden

The Selective Tuning model of visual attention and its impact on behaviour-based control architectures,

John Tsotsos

York University, Toronto, Canada

[157K]  Symbol-anchoring in Cassie

S.C. Shapiro and H.O. Ismail

Univ. of Buffalo, New York, USA

[140K]  Using behavior activation value histories for updating symbolic facts

F. Schönherr, M. Cistelecan, J. Hertzberg, and T. Christaller

Fraunhofer, Germany

[73K]  Toward bootstrap learning for place recognition

B.J. Kuipers and P. Beeson

Univ. of Texas, USA

[930K]  Perceptual attention through image feature generation based on visio-motor map learning

M. Asada and T. Minato

Osaka Univ., Japan

[108K]  From stereoscopic vision to symbolic representation

P.E. Santos and M.P. Shanahan

Imperial College, UK

[541K]  Interpreting symbols on conceptual spaces: the case of dynamic scenarios

A. Chella, M. Frixione, and S. Gaglio

Univ. of Palermo, Italy, and Univ. of Salerno, Italy

[251K]  Deriving fluents from sensor data for mobile robots

M. Witkowski, D. Randell, and M.P. Shanahan

Imperial College, UK

[231K]  Perception-action coupling via imitation and attention

G. Maistros, Y. Marom, and G. Hayes

Univ. of Edimburgh, UK

[2.0M]  Learning task representations from experienced demonstrations

M.N. Nicolescu and M.J. Mataric

Univ. of Southern California, USA

[123K]  Embodied AI: symbol grounding through imagination

H. Tsukimoto

Tokyo Denki Univ., Japan

[56K]  Some similarities between anchoring and pattern recognition concepts

I. Bloch and A. Saffiotti

ENST, France, and Orebro Univ., Sweden

[162K]  Anchoring: do we need new solutions to an old problem or do we have old solutions for a new problem?

A. Bonarini, M. Matteucci, and M. Restelli

Politecnico of Milano, Italy

[238K]  Symbol grounding in communicative mobile robots

P. Vogt

Univ. of Maastricht, The Netherlands

[157K]  HIVEMind: grounding interface in cooperative activity

A. Khoo and I.D. Horswill

Northwestern University, Illinois, USA

[85K]  Grounded models as a basis for intutitive reasoning: the origins of logical categories

J. Sierra-Santibáñez

Autonomous Univ. of Madrid, Spain

[48K]  Vision-based tracking for neutral buyoancy simulation

E.M. Atkins and R.S. Peasco

Univ. of Maryland, Maryland, USA

[90K]  Grounding robot behaviors

L. Hugues and A. Drogoul

Pierre & Marie Curie Univ., France

[14K]  Project Joshua Blue: common sense via common experience

S.S. Adams, S. Burbeck, N. Alvarado, and C. Latta

IBM Reseach, USA

[122K]  Grounded symbolic communication requires competitive adaptation of representations

D. Jung

Oak Ridge National Laboratory, USA

[57K]  Labeling for a learning mobile robot

N. Bredeche and J.D. Zucker

Pierre & Marie Curie Univ., France

[18K]  Towards memorizing by adjectives

A. Ayesh

De Montfor Univ., UK

[ Symposium Home ]

Website hosted by
Last updated on Oct 2, 2001