Part I. Chair : J. Droulez & J. Dokic

9h00 - 9h15 : J. Droulez (LPPA, CNRS), J. Dokic (CREA, Rouen University), N. Bullot (CREA, LPPA) Ouverture et prＴentation du colloque

9h15 - 10h00 : H. B〕thoff (Max-Planck-Institut f〉 Biologische Kybernetik, T｜ingen) Image-based object recognition

10h00 - 10h45 : P. Jacob (Institut de Sciences Cognitives, CNRS, Bron) Do size-contrast illusions deceive grasping ?

10h45 - 11h00 : Pause

11h00 - 11h45 : M. J. Tarr (Brown University) Understanding what the "face area" is computing through behavior, brain injury, and fMRI

11h45 - 12h30: K. O’Regan (Laboratoire de Psychologie ExpＳimentale, CNRS, Paris) Visual experience is not something we feel but something we do: change blindness as an empirical consequence of a sensorimotor theory of seeing.

12h30 - 14h00 : Pause

Part II. Chair : E. Pacherie (CREA, CNRS) & J. Lorenceau (IAF, CNRS)

14h00 - 14h45 : G. Rees (Institut of Neurology, University College London; California Institute of Technology) How does object perception depend on visual attention?

14h45 - 15h30 : S. Edelman (Cornell University) An effectively systematic exemplar-based framework for the representation of visual structure

15h30 - 16h00: Pause

16h00 - 16h45 : J.-L. Berm｣dez (University of Stirling) Objects, properties and two types of binding

16h45 – 17h00 : End

Organisation : N.J. Bullot, J. Dokic, J. Droulez (Atelier ‘La construction de l’objet per㎡’).

ABSTRACTS

Image-based object recognition

Heinrich B〕thoff

Max-Planck-Institute for Biological Cybernetics, T｜ingen, Germany

I will report on recognition experiments with (1) unfamiliar objects, (2) objects embedded in scenes, (3) familiar objects, (4) dynamic objects and (5) cross-modal transfer between visual and haptic recognition. All these experiments show strong viewpoint effects and speak in favor of an image-based representation of objects in which the physical similarity can account for recognition with small viewpoint changes. Recently, together with Guy Wallis we started to look at the importance of temporal similarity on the representation and recognition of objects. Temporal similarity can link many views to one object identify, because different views of objects are usually seen in close succession. To test this hypothesis subjects were presented sequences of novel faces in which the identity of the face changed as the head rotated. The subjects showed a tendency to treat the views as if they were of the same person. The results counter the proposal that object views are recognized simply on the basis of objective, structural components. Instead, they suggest that we are continuously associating views of objects to support later recognition, and that we do so not only on the basis of their physical similarity, but also their correlated appearance in time.

Do size-contrast illusions deceive grasping?

Pierre Jacob

Institut des sciences cognitives, CNRS, Bron, France

The goal of this paper is to assess the significance of recent experimental findings on some visual illusions - size-contrast illusions - for the dualistic model of the human visual system argued for by Goodale and Milner. According to their view, the human brain processes visual information in two fundamentally different ways according to whether its task is to provide a visual representation of the distal layout (visual food for thought, as it were), or to provide visual guidance for action and behavior upon the environment. We often act upon, or react to, objects in our environment without full visual awareness of them. Experienced drivers produce appropriate motor responses to visual stimuli of which they are dimly aware, if at all, when they act. Seeing can make us visually aware of objects, properties and facts in our environment. But it need not. Often, seeing allows us to act efficiently on objects of which we are not fully aware while we act. Primates and particularly humans are unique among animals in being able to grasp and manipulate objects in their environment using the dexterity of their hands. This is why in much of the literature on the so-called "visuomotor transformation", relevant actions are constituted by arm and hand motions directed towards objects, such as reaching and grasping. The present paper is mostly methodological. After sketching the rationale for Goodale and Milner's dualistic picture of the human visual system mostly based on anatomical work on the visual system of non-human primates and on brain-lesioned human patients, I examine several experiments dealing with visual size-contrast illusions. I will particularly focus on Titchener circles illusions. Visual illusion is a typical feature of normal conscious visual experience. Prima facie, the experiments reviewed here suggest that visual size-contrast illusions do not affect reaching and grasping as much as perceptual judgement. These experiments have been subject to some methodological criticisms which in turn have given rise to new experimental evidence. I will look both at the methodological criticisms and at the experimental findings. My tentative conclusion will be that the dualistic model of visual information processing survives the criticisms.

Understanding what the "face area" is computing through behavior, brain injury, and fMRI.

Michael J. Tarr

Brown University, USA

At least three approaches have been developed for understanding processing in category-specific areas in the human brain and most notably midFG, sometimes referred to as the "face area." First, psychophysical studies have revealed a range of apparently face-specific behavioral effects - in particular, extreme sensitivity to the configural aspects of stimuli across a variety of recognition tasks. Second, there are neuropsychological case studies where brain-injured subjects appear to be disproportionately impaired at face recognition as compared to the recognition of common non-face objects (prosopagnosia). Third, functional brain imaging studies (fMRI and PET) appear to show a dedicated neural substrate - a portion of inferior temporal lobe (IT; part of the fusiform and inferior temporal gyri) - in humans that is more active when viewing faces as compared to when viewing common non-face objects.

But what is midFG actually computing? It is our contention that engagement of this area is the result of within-category/item-specific recognition, in which members of a visually homogeneous class must be distinguished by experts, for instance, face recognition. Using this framework, we have examined

what are the specific conditions under which putatively face-specific behavioral effects are obtained, under which patients exhibit recognition deficits, and under which midFG is active. In behavioral studies, we found that experts, but not novices, with a novel class of non-face objects ("Greebles") showed configural sensitivity similar to that obtained with faces. Second, across multiple experiments we found that prosopagnosic subjects are disproportionally sensitive to increasingly specific levels of categorization with both familiar and novel non-face objects (rather than faces per se). Similarly, in one fMRI study we found that the brain region active in face processing is also during Greeble processing, but only following expertise training. Finally, we have explored whether this region is implicated more in the detection of faces or other overlearned categories, or rather, is part of a network of processes used for expert-level recognition of highly similar objects. Our results suggest that the level of visual categorization and the degree of expertise, not just stimulus class membership, are critical for understanding processing in the putative face area. Moreover, this brain region is simply one of several areas of visual cortex involved in the complex task of within-category recognition.

Visual experience is not something we feel but something we do: change blindness as an empirical consequence of a sensorimotor theory of seeing.

K. O’Regan

Laboratoire de Psychologie ExpＳimentale, CNRS, Paris, France

Any theory of visual experience which postulates that a brain mechanism generates the "raw feel" of seeing encounters the impassable "explanatory gap" separating physics from phenomenology. A way round the problem is to postulate that experience is not something we feel, but something we do: a kind of give-and-take with the environment analogous to the "feel" of driving a car. A consequence of this sensorimotor theory of visual experience is that the visual world is not re-presented in the brain, but acts like an outside memory store, accessible at the slightest flick of the eye or of attention. Though we have the impression of seeing everything in the visual field, only the very limited aspects of the world that we are attending to can be reported. The phenomenon of "change blindness" is one of a number of empirically testable consequences.

How does object perception depend on visual attention?

Geraint Rees

California Institute of Technology, USA

What we see is not dictated solely by where we point our eyes, but also by how we direct our visual attention. The effects of directed attention on visual object perception can be profound. While attending to a stimulus can substantially enhance a subject’s capacity to perform fine discriminations, an ignored stimulus sometimes appears not even to reach awareness. However, the nature and extent of processing for attended versus ignored stimuli has been a subject of controversy for many decades, and the brain mechanisms leading to such effects have until recently been unclear. This presentation will consider our recent investigations of the nature and extent of processing for unattended stimuli using functional magnetic resonance imaging (fMRI) in humans, studying both normal and brain damaged individuals.

Patients with disorders of attention due to focal cerebral lesions may fail to report awareness of visual stimuli presented in the contralesional visual field. Similarly, normal subjects may show reduced awareness for ignored visual stimuli when attention is fully occupied elsewhere. In a patient with visual extinction caused by focal right parietal damage, extinguished and unseen visual stimuli nevertheless evoked neural activity in contralateral striate and early extrastriate cortex. Activity in these structures is therefore not sufficient to produce awareness. In normal subjects, progressively engaging attention leads to reduced activity in visual areas processing irrelevant stimuli. Moreover, if attention is fully engaged elsewhere, then even highly familiar objects presented foveally fail to evoke brain activity related to their identity. Such 'inattentional blindness' shows that visual object recognition is wholly dependent on attention even for highly familiar and meaningful stimuli at the centre of gaze. Finally, evidence will be presented that links brain areas involved in two previously unrelated phenomena, bistable perception during binocular rivalry and visuospatial attention. This suggests that the mechanisms for switching between rival stimulus interpretations and between foci of attention share a common neural substrate, that plays a central role in the selection of neuronal events leading to visual awareness.

Taken together, these findings illustrate the strong relationship between object perception, visual awareness and selective attention, and show how brain imaging may illuminate some of the most enduring controversies in attention research.

An effectively systematic exemplar-based framework for the representation of visual structure

Shimon Edelman

Cornell University, USA

The problem of representing the spatial structure of images, which arises in visual object and scene processing, is commonly described using terminology borrowed from propositional theories of cognition, notably, the concept of compositionality. The classical propositional view mandates representations composed of symbols, which stand for atomic or composite entities, and enter into arbitrarily nested relationships. We argue that the main desiderata of a representational system – productivity and systematicity – can be achieved without recourse to the classical, proposition-like compositionality. We outline a systematic and productive framework for the representation of visual structure, which relies on shallow rather than nested compositionality and can use shape primitives that are coarsely coded rather than atomic, and concrete rather than generic.

Joint work with Nathan Intrator.

Objects, properties and two types of binding

Jos� Luis Berm｣dez

University of Stirling, United Kingdom

Researchers into object perception have placed considerable emphasis on solving the so-called property binding problem. The property binding problem, as it is usually presented (e.g. Treisman 1996), arises because different aspects of an object are separately processed in specialised visual areas. A solution to the property binding problem will be an explanation of how a set of properties are bound to each other as properties of a single object. Many discussions of the binding problem do not take into account, however, the distinction between perceiving a collection of co-instantiated properties and perceiving an object. Object perception involves more than the perception of a cluster of features located at a single place (Strawson 1959). This creates a further binding problem. Even once we've explained what it is for a set of properties to be experienced as co-instantiated we still have to explain what it is for them to be experienced as properties of a single object.

This paper rejects the classical metaphysical approach that characterises objects as bundles of properties that have a common bearer, in favour of individuating objects as bundles of features that obey certain high-level physical principles. This has obvious implications for the binding problem. What makes it the case that a co-instantiated set of properties is perceived as a single object is that the bundle is perceived as obeying the relevant high-level physical principles. This has the further consequence, however, that we need to recognise the existence of forms of perceptual sensitivity to objects that do not count as genuine forms of object perception - because they do not involve perception of the right set of physical principles. The paper explores this possibility in the context of current research on object permanence in infancy.