A model of visual recognition and categorization
Philosophical Transactions of the Royal Society B Biological SciencesPublished 29 August 1997Open access
Shimon Edelman, Sharon Duvdevani-Bar
Citations112
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
The resulting representational scheme based on similarities to prototypes appears to be computationally viable, and is readily mapped onto the mechanisms of biological vision revealed by recent psychophysical and physiological studies.
Abstract
To recognize a previously seen object, the visual system must overcome the variability in the object's appearance caused by factors such as illumination and pose. Developments in computer vision suggest that it may be possible to counter the influence of these factors, by learning to interpolate between stored views of the target object, taken under representative combinations of viewing conditions. Daily life situations, however, typically require categorization, rather than recognition, of objects. Due to the open-ended character of both natural and artificial categories, categorization cannot rely on interpolation between stored examples. Nonetheless, knowledge of several representative members, or prototypes, of each of the categories of interest can still provide the necessary computational substrate for the categorization of new instances. The resulting representational scheme based on similarities to prototypes appears to be computationally viable, and is readily mapped onto the mechanisms of biological vision revealed by recent psychophysical and physiological studies.
Keywords
Neuroscience
IEEE Transactions on Information TheoryNearest neighbor pattern classification
16,134 Citations1967Thomas M. Cover, Peter E. Hart
The nearest neighbor decision rule assigns to an unclassified sample point the classification of the nearest of a set of previously classified points, so it may be said that half the classification information in an infinite sample set is contained in the nearest neighbor.
Pattern classification and scene analysis
12,643 Citations1973Richard O. Duda, Peter E. Hart
IEEE Transactions on CommunicationsAn Algorithm for Vector Quantizer Design
7,183 Citations1980Y. Linde, A. Buzo +1 more
An efficient and intuitive algorithm is presented for the design of vector quantizers based either on a known probabilistic model or on a long training sequence of data.
Psychological ReviewRecognition-by-components: A theory of human image understanding.
5,507 Citations1987Irving Biederman
Recognition-by-components (RBC) provides a principled account of the heretofore undecided relation between the classic principles of perceptual organization and pattern recognition.
Complex SystemsRadial Basis Functions, Multi-Variable Functional Interpolation and Adaptive Networks
3,439 Citations1988David S. Broomhead, David Lowe
The relationship between 'learning' in adaptive layered networks and the fitting of data with high dimensional surfaces is discussed, leading naturally to a picture of 'generalization in terms of interpolation between known data points and suggests a rational approach to the theory of such networks.
Proceedings of the Royal Society of London. Series B, Biological sciencesRepresentation and recognition of the spatial organization of three-dimensional shapes
2,183 Citations1978David Marr, H. K. Nishihara
The human visual process can be studied by examining the computational problems associated with deriving useful information from retinal images by applying the approach to the problem of representing three-dimensional shapes for the purpose of recognition.
International Journal of Computer VisionVisual learning and recognition of 3-d objects from appearance
1,891 Citations1995Hiroshi Murase, Shree K. Nayar
A near real-time recognition system with 20 complex objects in the database has been developed and a compact representation of object appearance is proposed that is parametrized by pose and illumination.
Bulletin of the London Mathematical SocietyShape Manifolds, Procrustean Metrics, and Complex Projective Spaces
1,478 Citations1984David G. Kendall
Annual Review of NeuroscienceInferotemporal Cortex and Object Vision
1,413 Citations1996Keiji Tanaka
Cells in area TE of the inferotemporal cortex of the monkey brain selectively respond to various moderately complex object features, and those that cluster in a columnar region that runs perpendicular to the cortical surface respond to similar features.
Principles of Categorization
1,152 Citations2024Eleanor Rosch
ScienceRegularization Algorithms for Learning That Are Equivalent to Multilayer Networks
1,042 Citations1990Tomaso Poggio, Federico Girosi
A theory is reported that shows the equivalence between regularization and a class of three-layer networks called regularization networks or hyper basis functions.
Current BiologyShape representation in the inferior temporal cortex of monkeys
980 Citations1995Nikos K. Logothetis, J Pauls +1 more
The results suggest that IT neurons can develop a complex receptive field organization as a consequence of extensive training in the discrimination and recognition of objects, and support the idea that a population of neurons tuned to a different object aspect, and each showing a certain degree of invariance to image transformations--may encode at least some types of complex three-dimensional objects.
NatureA network that learns to recognize three-dimensional objects
936 Citations1990Tomaso Poggio, Shimon Edelman
A scheme is developed, based on the theory of approximation of multivariate functions, that learns from a small set of perspective views a function mapping any viewpoint to a standard view, and a network equivalent to this scheme will 'recognize' the object on which it was trained from any viewpoint.
IEEE Transactions on Pattern Analysis and Machine IntelligenceRecognition by linear combinations of models
817 Citations1991Shimon Ullman, Ronen Basri
An approach to visual object recognition in which a 3D object is represented by the linear combination of 2D images of the object is proposed and it is shown that for objects with sharp edges as well as with smooth bounding contours, the set of possible images of a given object is embedded in a linear space spanned by a small number of views.
ScienceMultidimensional Scaling, Tree-Fitting, and Clustering
754 Citations1980Roger N. Shepard
Applications to perceptual and semantic data illustrate how complementary aspects of the underlying psychological structure are revealed by different types of representations, including multidimensional spatial configurations and nondimensional tree-structures or clusterings.
Proceedings of the National Academy of SciencesPsychophysical support for a two-dimensional view interpolation theory of object recognition.
682 Citations1992HH Bülthoff, Shimon Edelman
The results suggest that the human visual system is better described as recognizing these objects by two-dimensional view interpolation than by alignment or other methods that rely on object-centered three-dimensional models.
Cognitive PsychologySurface versus edge-based determinants of visual recognition
656 Citations1988Irving Biederman, Ginny Ju
Although differences in surface characteristics such as color, brightness, and texture can be instrumental in defining edges and can provide cues for visual search, they play only a secondary role in the real-time recognition of an intact object when its edges can be readily extracted.
CognitionAligning pictorial descriptions: An approach to object recognition
572 Citations1989Shimon Ullman
The proposed alignment method also uses abstract description, but unlike structural description methods it uses them pictorially, rather than in symbolic structural descriptions.
Cognitive PsychologyPictures and names: Making the connection
542 Citations1984Pierre Jolicœur, Mark A. Gluck +1 more
A series of results supports two general assertions about this process: first, objects are identified first at a particular level of abstraction which is neither the most general nor the most specific possible, and second, the particular entry point for a given object covaries with its typicality, which affects whether or not the object will be identified at the “basic” level.
The MIT Press eBooksHigh-Level Vision
429 Citations1996Shimon Ullman
Object recognition: shape-based recognition what is recognition, why object recognition is difficult, and approaches to object recognition the alignment approach which is the correct approach?
Neural ComputationSEEMORE: Combining Color, Shape, and Texture Histogramming in a Neurally Inspired Approach to Visual Object Recognition
344 Citations1997Bartlett W. Mel
It is concluded that in the light of the vast hardware resources available in the ventral stream of the primate visual system relative to those exercised here, the appealingly simple feature-space conjecture remains worthy of serious consideration as a neurobiological model.
The Quarterly Journal of Experimental Psychology Section AThe Effects of Surface Detail on Object Categorization and Naming
306 Citations1989Cathy J. Price, Glyn W. Humphreys
The results suggest that surface details can affect object recognition and naming, depending upon: (1) the degree to which objects must be differentiated for a correct response to be made, and (2) the nature of the rate-limiting process determining performance.
ScienceOptical Imaging of Functional Organization in the Monkey Inferotemporal Cortex
305 Citations1996Gang Wang, Keiji Tanaka +1 more
To investigate the functional organization of object recognition, the technique of optical imaging was applied to the primate inferotemporal cortex, which is thought to be essential for object recognition.
Lecture notes in computer scienceObject recognition using multidimensional receptive field histograms
238 Citations1996Bernt Schiele, James L. Crowley
The mathematical foundations of the technique are described and the results of experiments which compare robustness and recognition rates for different local neighborhood operators and histogram similarity measurements are presented.
Minds and MachinesRepresentation, similarity, and the chorus of prototypes
164 Citations1995Shimon Edelman
The present preliminary study appears to indicate that this concept of representation is computationally viable, and is compatible with psychological and neurobiological data.
Current Opinion in NeurobiologyInferotemporal cortex and higher visual functions
87 Citations1992Keiji Tanaka
A columnar organization has been revealed, and a short-term change of responsiveness has been proposed as a neural correlate of working memory.
Cognitive PsychologyPerceptual-cognitive explorations of a toroidal set of free-form stimuli
83 Citations1973Roger N. Shepard, Gregory W. Cermak
The Psychology of learning and motivation/The psychology of learning and motivationLearning as Extraction of Low-Dimensional Representations
66 Citations1997Shimon Edelman, Nathan Intrator
This chapter discusses that the effectiveness of living representational systems suggest that there must be something special about such systems that allows them to harbor representations of the world, and suggests that the possibility of effective representation stems from theLow-dimensional nature of the real-world classification tasks: an intelligent system would do well merely by reflecting the low-dimensional distal space internally.
Proceedings of the National Academy of SciencesFaithful representation of similarities among three-dimensional shapes in human vision.
53 Citations1996Florin Cutzu, Shimon Edelman
This work investigated the ability of the human visual system to derive representations from the sensory input-a highly nontrivial task, given the million or so dimensions of the visual signal at its entry point to the cortex.
Neural ComputationRepresentation of Similarity in Three-Dimensional Object Discrimination
51 Citations1995Shimon Edelman
The results of four tests support the notion that discrimination among 3D objects may rely on a low-dimensional feature space representation, and suggest that this space may be spanned by explicitly encoded class prototypes.
Lecture notes in computer scienceLearning to recognize faces from examples
50 Citations1992Shimon Edelman, D. Reisfeld +1 more
An implemented system that learns to recognize human faces under varying pose and illumination conditions that relies on symmetry operations to detect the eyes and the mouth in a face image, and performs simple but effective dimensionality reduction by a convolution.
International Journal of Computer VisionRecognition by prototypes
41 Citations1996Ronen Basri
The paper concludes with an algorithm for constructing optimal prototypes for classes of objects, which combines the correspondence and object pose computed in the categorization stage to align the prototype with the image with the result that identification is reduced to a series of simple template comparisons.
Neural ComputationSimilarity, Connectionism, and the Problem of Representation in Vision
40 Citations1997Shimon Edelman, Sharon Duvdevani-Bar
The computational requirements of representation that preserves similarity ranks are discussed and the relative straightforwardness of its connectionist implementation is pointed out.
CogPrints (Cogprints)Representation of Similarity in 3D Object Discrimination
27 Citations1995Shimon Edelman
IEEE Transactions on Pattern Analysis and Machine IntelligenceThe space requirements of indexing under perspective projections
21 Citations1996David W. Jacobs
It is shown that in this case, one must use a 3D surface to represent model groups, and an analytic description of such a surface is provided, in contrast to the cases of scaled-orthographic or affine projection.
A pictorial approach to object classification
9 Citations1991Yerucham Shapira, Shimon Ullman
This work extends previous alignment schemes in two main directions: extending the set of allowed transformations between the model and the viewed object, and using structural aspects of the internal models, namely, their part decomposition.
CogPrints (Cogprints)Receptive Fields for Vision: from Hyperacuity to Object Recognition
8 Citations1995Shimon Edelman
This paper surveys some of the possible computational reasons behind the ubiquity of RFs, by discussing examples of RF-based solutions to problems in vision, from spatial acuity, through sensory coding, to object recognition.
Neuroscience Research2501 Representation of similarity among 3D shapes in the monkey visual system
1 Citations1997Tadashi Sugihara, Shimon Edelman +1 more
The possibility of veridical representation of parametric similarity among complex objects in the monkey visual system is demonstrated and the MDS solution resembled closely the 2D triangle built into the stimuli.