Attention in Cognitive Systems: 5th International Workshop on Attention in Cognitive Systems, WAPCV 2008 Fira, Santorini, Greece, May 12, 2008 Revised Selected Papers PDF
Preview Attention in Cognitive Systems: 5th International Workshop on Attention in Cognitive Systems, WAPCV 2008 Fira, Santorini, Greece, May 12, 2008 Revised Selected Papers
Lecture Notes in Artificial Intelligence 5395 EditedbyR.Goebel,J.Siekmann,andW.Wahlster Subseries of Lecture Notes in Computer Science Lucas Paletta John K. Tsotsos (Eds.) Attention in Cognitive Systems 5th International Workshop onAttentioninCognitiveSystems,WAPCV2008 Fira, Santorini, Greece, May 12, 2008 Revised Selected Papers 1 3 SeriesEditors RandyGoebel,UniversityofAlberta,Edmonton,Canada JörgSiekmann,UniversityofSaarland,Saarbrücken,Germany WolfgangWahlster,DFKIandUniversityofSaarland,Saarbrücken,Germany VolumeEditors LucasPaletta JoanneumResearch InstituteofDigitalImageProcessing Wastiangasse6,8010Graz,Austria E-mail:[email protected] JohnK.Tsotsos YorkUniversity CenterforVisionResearch(CVR) andDepartmentofComputerScienceandEngineering 4700KeeleSt.,TorontoONM3J1P3,Canada E-mail:[email protected] LibraryofCongressControlNumber:2009921734 CRSubjectClassification(1998):I.2,I.4,I.5,I.3,J.3 LNCSSublibrary:SL7–ArtificialIntelligence ISSN 0302-9743 ISBN-10 3-642-00581-0SpringerBerlinHeidelbergNewYork ISBN-13 978-3-642-00581-7SpringerBerlinHeidelbergNewYork Thisworkissubjecttocopyright.Allrightsarereserved,whetherthewholeorpartofthematerialis concerned,specificallytherightsoftranslation,reprinting,re-useofillustrations,recitation,broadcasting, reproductiononmicrofilmsorinanyotherway,andstorageindatabanks.Duplicationofthispublication orpartsthereofispermittedonlyundertheprovisionsoftheGermanCopyrightLawofSeptember9,1965, initscurrentversion,andpermissionforusemustalwaysbeobtainedfromSpringer.Violationsareliable toprosecutionundertheGermanCopyrightLaw. springer.com ©Springer-VerlagBerlinHeidelberg2009 PrintedinGermany Typesetting:Camera-readybyauthor,dataconversionbyScientificPublishingServices,Chennai,India Printedonacid-freepaper SPIN:12633487 06/3180 543210 Preface Attention has represented a core scientific topic in the design of AI-enabled systems in the last few decades.Today,in the ongoingdebate, design,andcom- putationalmodelingofartificialcognitivesystems,attentionhasgainedacentral positionasa focus ofresearch.For instance,attentionalmethods areconsidered in investigating the interfacing of sensory and cognitive information processing, for the organization of behaviors, and for the understanding of individual and social cognition in infant development. Whilevisualcognitionplaysacentralroleinhumanperception,findingsfrom neuroscience and experimental psychologyhave provided strong evidence about the perception–action nature of cognition. The embodied nature of sensory- motor intelligence requires a continuous and focused interplay between the con- trolofmotoractivitiesandtheinterpretationoffeedbackfromperceptualmodal- ities. Decision making about the selection of information from the incoming sensory stream – in tune with contextual processing on a current task and an agent’s global objectives – becomes a further challenging issue in attentional control.Attentionmustoperateatinterfacesbetweenabottom-up-drivenworld interpretationandtop-down-driveninformationselection,thusactingatthecore of artificial cognitive systems. These insights have already induced changes in AI-related disciplines, such as the design of behavior-based robot control and the computational modeling of animats. Today,thedevelopmentofenablingtechnologiessuchasautonomousrobotic systems,miniaturizedmobile–evenwearable–sensors,andambientintelligence systems involves the real-time analysis of enormous quantities of data. These data have to be processed in an intelligent way to provide “on time delivery” of the required relevant information. Knowledge has to be applied about what needs to be attended to, and when, and what to do in a meaningful sequence, in correspondence with visual feedback. Theindividualcontributionsofthisbookmeetthesescientificandtechnolog- icalchallengesonthedesignofattentionandpresentthelateststateoftheartin relatedfields.Thebookevolvedoutofthe5thInternationalWorkshoponAtten- tioninCognitiveSystems(WAPCV2008)thatwasheldonSantorini,Greece,as anassociatedworkshopofthe6thInternationalConferenceonComputerVision Systems (ICVS 2008). The goal of this workshop was to provide an interdisci- plinaryforumtoexaminecomputationalmodelsofattentionincognitivesystems from an interdisciplinary viewpoint, with a focus on computer visionin relation topsychology,roboticsandneuroscience.Theworkshopwasheldasasingle-day, single-track event, consisting of high-quality podium and poster presentations. Wereceivedatotalof34papersubmissionsforreview,22ofwhichwereretained for presentations (13 oral presentations and 9 posters). We would like to thank the members of the Program Committee for their substantial contribution to VI Preface the quality of the workshop. Two invited speakers strongly supported the suc- cessoftheeventwithwell-attendedpresentationsgivenon“LearningtoAttend: From Bottom-Up to Top-Down” (Jochen Triesch) and “Brain Mechanisms of Attentional Control” (Steve Yantis). WAPCV2008andtheeditingofthiscollectionwassupportedinpartbyThe EuropeanNetworkfortheAdvancementofArtificialCognitiveSystems(euCog- nition).WeareverythankfultoDavidVernon(co-ordinatorofeuCognition)and Colette Maloney of the European Commission’s ICT Programon Cognition for their financial and moral support. Finally, we wish to thank Katrin Amlacher for her efforts in assembling these proceedings. January 2009 Lucas Paletta John K. Tsotsos Organization Chairing Committee Lucas Paletta Joanneum Research, Austria John K. Tsotsos York University, Canada Advisory Committee Laurent Itti University of Southern California, CA (USA) Jan-Olof Eklundh KTH (Sweden) Program Committee Leonardo Chelazzi University of Verona, Italy James J. Clark McGill University, Canada J.M. Findlay Durham University, UK Simone Frintrop University of Bonn, Germany Fred Hamker University of Muenster, Germany Dietmar Heinke University of Birmingham, UK Laurent Itti University of Southern California, USA Christof Koch California Institute of Technology, USA Ilona Kovacs Budapest University of Technology, Hungary Eileen Kowler Rutgers University, USA Michael Lindenbaum Technion, Israel Larry Manevitz University of Haifa, Israel Baerbel Martsching University of Paderborn, Germany Giorgio Metta University of Genoa, Italy Vidhay Navalpakkam California Institute of Technology, USA Aude Oliva MIT, USA Kevin O’Regan Universit´e de Paris 5, France Fiora Pirri University of Rome, La Sapienza, Italy Marc Pomplun University of Massachusetts, USA Catherine Reed University of Denver, USA Ronald A. Rensink University of British Columbia, Canada Erich Rome Fraunhofer IAIS, Germany John G. Taylor King’s College London, UK Jochen Triesch Frankfurt Institute for Advanced Studies, Germany Nuno Vasconcelos University of California San Diego, USA Chen Yu University of Indiana, USA Tom Ziemke University of Skovde, Sweden VIII Organization Sponsoring Institutions euCognition - The European Network for the Advancement of Artificial Cogni- tive Systems Joanneum Research, Austria Table of Contents Attention in Scene Exploration On the Optimality of Spatial Attention for Object Detection .......... 1 Jonathan Harel and Christof Koch Decoding What People See from Where They Look: Predicting Visual Stimuli from Scanpaths ........................................... 15 Moran Cerf, Jonathan Harel, Alex Huth, Wolfgang Einha¨user, and Christof Koch A Novel Hierarchical Framework for Object-Based Visual Attention .... 27 Rebecca Marfil, Antonio Bandera, Juan Antonio Rodr´ıguez, and Francisco Sandoval Where Do We Grasp Objects? – An Experimental Verification of the Selective Attention for Action Model (SAAM) ....................... 41 Christoph Bo¨hme and Dietmar Heinke Contextual Cueing and Saliency Integrating Visual Contextand Object Detection within a Probabilistic Framework...................................................... 54 Roland Perko, Christian Wojek, Bernt Schiele, and Aleˇs Leonardis The Time Course of Attentional Guidance in Contextual Cueing ....... 69 Andrea Schankin and Anna Schub¨o Conspicuity and Congruity in Change Detection ..................... 85 Jean Underwood, Emma Templeman, and Geoffrey Underwood Spatiotemporal Saliency Spatiotemporal Saliency: Towards a Hierarchical Representation of Visual Saliency .................................................. 98 Neil D.B. Bruce and John K. Tsotsos Motion Saliency Maps from Spatiotemporal Filtering ................. 112 Anna Belardinelli, Fiora Pirri, and Andrea Carbone X Table of Contents Attentional Networks Model Based Analysis of fMRI-Data: Applying the sSoTS Framework to the Neural Basic of Preview Search .............................. 124 Eirini Mavritsaki, Harriet Allen, and Glyn Humphreys Modelling the Efficiencies and Interactions of Attentional Networks..... 139 Fehmida Hussain and Sharon Wood The JAMF Attention Modelling Framework ......................... 153 Johannes Steger, Niklas Wilming, Felix Wolfsteller, Nicolas Ho¨ning, and Peter K¨onig Attentional Modeling Modeling Attention and Perceptual Grouping to Salient Objects ....... 166 Thomas Geerinck, Hichem Sahli, David Henderickx, Iris Vanhamel, and Valentin Enescu Attention Mechanisms in the CHREST Cognitive Architecture......... 183 Peter C.R. Lane, Fernand Gobet, and Richard Ll. Smith Modeling the Interactions of Bottom-Up and Top-Down Guidance in Visual Attention ................................................. 197 David Henderickx, Kathleen Maetens, Thomas Geerinck, and Eric Soetens Relative Influence of Bottom-Up and Top-Down Attention ............ 212 Matei Mancas TowardsStandardizationof EvaluationMetrics and Methods for Visual Attention Models ................................................ 227 Muhammad Zaheer Aziz and Ba¨rbel Mertsching Comparing Learning Attention Control in Perceptual and Decision Space .......................................................... 242 Maryam S. Mirian, Majid Nili Ahmadabadi, Babak N. Araabi, and Ronald R. Siegwart Automated Visual Attention Manipulation .......................... 257 Tibor Bosse, Rianne van Lambalgen, Peter-Paul van Maanen, and Jan Treur Author Index.................................................. 273 On the Optimality of Spatial Attention for Object Detection Jonathan Harel and Christof Koch California Instituteof Technology, Pasadena, CA, 91125 Abstract. Studiesonvisualattentiontraditionally focusonitsphysio- logical and psychophysical nature [16,18,19], or its algorithmic applica- tions [1,9,21]. We here develop a simple, formal mathematical model of the advantage of spatial attention for object detection, in which spatial attention is defined as processing a subset of the visual input, and de- tection isan abstraction with certain failure characteristics. Wedemon- stratethatitissuboptimaltoprocesstheentirevisualinputgivenprior information about target locations, which in practice is almost always available in a video setting due to tracking, motion, or saliency. This arguesforanattentionalstrategyindependentofcomputationalsavings: no matter how much computational power is available, it is in principle bettertodedicateitpreferentiallytoselectedportionsofthescene.This suggests,anecdotally,aformofenvironmentalpressurefortheevolution of foveated photoreceptor densities in theretina. It also offers a general justification for the useof spatial attention in machine vision. 1 Introduction Most animals with visual systems have evolved the peculiar trait of processing subsets of the visual input at higher bandwidth (faster reaction times, lower error rates, higher SNR). This strategy is known as focal or spatial attention and is closely linked to sensory (receptor distribution in the retina) and motor (eyemovements)factors.Motivatedbysuchwide-spreadattentionalprocessing, many machine vision scientists have developed computational models of visual attention, with some treating it broadly as a hierarchical narrowing of possibil- ities [1,2,8,9,17]. Several studies have demonstrated experimental paradigms in whichvarioussuchattentionalschemesarecombinedwithrecognition/detection algorithms, and have documented the resulting computational savings and/or improved accuracy [4,5,6,7,20,21]. Here, we seek to describe a general justification for spatial attention in the context of an object detection goal (detecting targets in images wherever they occur). We take an abstract approach to this phenomenon, in which both the attentional and detection mechanisms are independent of the conclusions. Sim- ilar frameworks have been proposed by other authors [3,10]. The most common justification for attentional processing, in particular in visual psychology, is the computational saving that accrue if processing is restricted to a subset of the image.Formachine visionscientists, inanage ofever decreasingcomputational L.PalettaandJ.K.Tsotsos(Eds.):WAPCV2008,LNAI5395,pp.1–14,2009. (cid:1)c Springer-VerlagBerlinHeidelberg2009
The list of books you might like
