Computation of pattern invariance in brain-like structures

Ullman, Shimon; Soloviev, Sergei

doi:10.1016/s0893-6080(99)00048-9

Cited by 48 publications

(32 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Against this background, our findings add a novel perspective, as they demonstrate that invariance to positional changes is also a by-product of the top-down structuring of the visual world imposed by the process of category acquisition. In this way, position invariance induced by category learning might act in a complementary way to invariance mechanisms of more limited scope, which may be active at early and intermediate levels of feature processing and result from a conjunctive sampling of the visual field (Riesenhuber & Poggio 1999) or partial generalizations built upon past sensory experience (Ullman & Soloviev 1999). …”

Section: Discussionmentioning

confidence: 99%

Category learning induces position invariance of pattern recognition across the visual field

Jüttner

Rentschler

2007

Proc. R. Soc. B.

View full text Add to dashboard Cite

Human object recognition is considered to be largely invariant to translation across the visual field. However, the origin of this invariance to positional changes has remained elusive, since numerous studies found that the ability to discriminate between visual patterns develops in a largely location-specific manner, with only a limited transfer to novel visual field positions. In order to reconcile these contradicting observations, we traced the acquisition of categories of unfamiliar grey-level patterns within an interleaved learning and testing paradigm that involved either the same or different retinal locations. Our results show that position invariance is an emergent property of category learning. Pattern categories acquired over several hours at a fixed location in either the peripheral or central visual field gradually become accessible at new locations without any position-specific feedback. Furthermore, categories of novel patterns presented in the left hemifield are distinctly faster learnt and better generalized to other locations than those learnt in the right hemifield. Our results suggest that during learning initially position-specific representations of categories based on spatial pattern structure become encoded in a relational, position-invariant format. Such representational shifts may provide a generic mechanism to achieve perceptual invariance in object recognition.

show abstract

Section: Discussionmentioning

confidence: 99%

Category learning induces position invariance of pattern recognition across the visual field

Jüttner

Rentschler

2007

Proc. R. Soc. B.

View full text Add to dashboard Cite

show abstract

“…This indicates that, in early trials, rat performance was mainly accounted for by the degree of spontaneously perceived similarity between the novel and the default prototype appearances, while, during the course of training, a fuller tolerance was gradually achieved by explicitly learning the associative relations among the different appearances of each prototype (Miyashita, 1993). This suggests that also for rats, as proposed for primates (Logothetis et al, 1994;Bülthoff et al, 1995;Tarr and Bülthoff, 1998;Lawson, 1999;Afraz and Cavanagh, 2008;Kravitz et al, 2008Kravitz et al, , 2010 and successfully implemented in many leading artificial vision systems (Poggio and Edelman, 1990;Riesenhuber and Poggio, 1999;Ullman and Soloviev, 1999;Ullman, 2007), transformationtolerant recognition is achieved by combining the limited (but automatic) tolerance granted by banks of partially tolerant feature detectors with the fuller tolerance obtained by interpolating between stored representations of multiple, independently learned object views.…”

Section: Validity and Implications Of Our Findingsmentioning

confidence: 91%

“…Therefore, priming and adaptation aftereffect studies can disentangle the component of transformationtolerant recognition that relies on spontaneously perceiving as similar different appearances of an object from the contribution of explicitly learning the associative relations among such object appearances. Mechanistically, this provides useful insight into the capability of visual object representations to support generalization of recognition to fully novel, never-before-experienced object appearances, which is the major computational feat that any biological or artificial recognition system has to face (Ullman and Soloviev, 1999;Riesenhuber and Poggio, 2000;Ullman, 2000). As an example, two recent studies (Afraz and Cavanagh, 2008;Kravitz et al, 2010) exploited adaptation and aftereffect paradigms to show that translation tolerance of face and object representations in human visual cortex is far more limited than commonly assumed.…”

Section: Validity and Implications Of Our Findingsmentioning

confidence: 99%

Transformation-Tolerant Object Recognition in Rats Revealed by Visual Priming

Tafazoli

Filippo²,

Zoccolan³

2012

J. Neurosci.

View full text Add to dashboard Cite

Successful use of rodents as models for studying object vision crucially depends on the ability of their visual system to construct representations of visual objects that tolerate (i.e., remain relatively unchanged with respect to) the tremendous changes in object appearance produced, for instance, by size and viewpoint variation. Whether this is the case is still controversial, despite some recent demonstration of transformation-tolerant object recognition in rats. In fact, it remains unknown to what extent such a tolerant recognition has a spontaneous, perceptual basis, or, alternatively, mainly reflects learning of arbitrary associative relations among trained object appearances. In this study, we addressed this question by training rats to categorize a continuum of morph objects resulting from blending two object prototypes. The resulting psychometric curve (reporting the proportion of responses to one prototype along the morph line) served as a reference when, in a second phase of the experiment, either prototype was briefly presented as a prime, immediately before a test morph object. The resulting shift of the psychometric curve showed that recognition became biased toward the identity of the prime. Critically, this bias was observed also when the primes were transformed along a variety of dimensions (i.e., size, position, viewpoint, and their combination) that the animals had never experienced before. These results indicate that rats spontaneously perceive different views/appearances of an object as similar (i.e., as instances of the same object) and argue for the existence of neuronal substrates underlying formation of transformation-tolerant object representations in rats.

show abstract

“…Complicating matters, the relationship between invariance and conjunction selectivity can depend on a particular measure of activity that is used as proxy to quantify the complexity of features that drive each neuron (1, 3, 10). This raises the question as to what neural architectures can ultimately sustain reliable object recognition, a question that is also currently at the forefront of computer vision (15,16).To provide constraints helpful in addressing this question we focused on the area V4, an intermediate area within the visual object recognition pathway that collects signals from areas V1 and V2 and provides input to the inferotemporal cortex. V4 neurons have previously been shown to be selective to curvature (17)(18)(19)(20).…”

mentioning

confidence: 99%

mentioning

confidence: 99%

Trade-off between curvature tuning and position invariance in visual area V4

Sharpee

Kouh

Reynolds

2013

Proc. Natl. Acad. Sci. U.S.A.

View full text Add to dashboard Cite

Humans can rapidly recognize a multitude of objects despite differences in their appearance. The neural mechanisms that endow high-level sensory neurons with both selectivity to complex stimulus features and "tolerance" or invariance to identity-preserving transformations, such as spatial translation, remain poorly understood. Previous studies have demonstrated that both tolerance and selectivity to conjunctions of features are increased at successive stages of the ventral visual stream that mediates visual recognition. Within a given area, such as visual area V4 or the inferotemporal cortex, tolerance has been found to be inversely related to the sparseness of neural responses, which in turn was positively correlated with conjunction selectivity. However, the direct relationship between tolerance and conjunction selectivity has been difficult to establish, with different studies reporting either an inverse or no significant relationship. To resolve this, we measured V4 responses to natural scenes, and using recently developed statistical techniques, we estimated both the relevant stimulus features and the range of translation invariance for each neuron. Focusing the analysis on tuning to curvature, a tractable example of conjunction selectivity, we found that neurons that were tuned to more curved contours had smaller ranges of position invariance and produced sparser responses to natural stimuli. These trade-offs provide empirical support for recent theories of how the visual system estimates 3D shapes from shading and texture flows, as well as the tiling hypothesis of the visual space for different curvature values.A lthough object recognition feels effortless, it is in fact a challenging computational problem (1). There are two important properties that any system that mediates robust object recognition must have. The first property is known as "invariance": the ability of the system to respond similarly to different views of the same object. The second property is known as "selectivity." Selectivity requires that systems' components, such as neurons within the ventral visual stream, produce different responses to potentially quite similar objects (such as different faces) even when presented from similar viewpoints. It is straightforward to make detectors that are invariant but not selective or selective but not invariant. The difficulty lies in how to make detectors that are both selective and invariant.To address this problem, both computer object recognition algorithms (2) and neural systems use a series of hierarchical stimulus representations, increasing both in complexity and the range of invariance (1, 3). For example, in each successive area of visual processing, neurons become selective for increasingly complex stimulus features (4-9) and grow more tolerant to identity-preserving transformations, such as image translation, scaling, and, to some degree, rotation and the presence of "clutter" from other objects in the scene (3,(10)(11)(12). This has led to the idea that high-level sensory neurons are...

show abstract

Computation of pattern invariance in brain-like structures

Cited by 48 publications

References 27 publications

Category learning induces position invariance of pattern recognition across the visual field

Category learning induces position invariance of pattern recognition across the visual field

Transformation-Tolerant Object Recognition in Rats Revealed by Visual Priming

Trade-off between curvature tuning and position invariance in visual area V4

Contact Info

Product

Resources

About