Encoding multielement scenes: statistical learning of visual feature hierarchies.

@article{Fiser2005EncodingMS,
  title={Encoding multielement scenes: statistical learning of visual feature hierarchies.},
  author={J{\'o}zsef Fiser and Richard N. Aslin},
  journal={Journal of experimental psychology. General},
  year={2005},
  volume={134 4},
  pages={
          521-37
        }
}
  • J. Fiser, R. Aslin
  • Published 1 November 2005
  • Computer Science
  • Journal of experimental psychology. General
The authors investigated how human adults encode and remember parts of multielement scenes composed of recursively embedded visual shape combinations. The authors found that shape combinations that are parts of larger configurations are less well remembered than shape combinations of the same kind that are not embedded. Combined with basic mechanisms of statistical learning, this embeddedness constraint enables the development of complex new features for acquiring internal representations… 
View on PubMed
bcs.rochester.edu
181 Citations
Highly Influential Citations
15
Background Citations
89
Methods Citations
10
Results Citations
9

Figures from this paper

181 Citations

Citation Type

Citation Type

Parts beget parts: Bootstrapping hierarchical object representations through visual statistical learning
The role of Gestalt grouping principles in visual statistical learning
TLDR
It is shown that perceptual grouping cues are also incorporated in this Bayesian model, providing additional evidence for the possible span of chunks, and that the very same embedded chunks are learned if they are grouped by perceptual cues.
Bayesian learning of visual chunks by human observers
TLDR
An ideal learner based on Bayesian model comparison that extracts and stores only those chunks of information that are minimally sufficient to encode a set of visual scenes is developed.
Visual learning of statistical relations among nonadjacent features: Evidence for structural encoding
TLDR
Observers exhibited learning consistent with earlier studies, suggesting that unsupervised learning of compositional structure is based on the explicit encoding of spatial relations between separable visual features.
Neural representations for object perception: structure, category, and adaptive coding.
TLDR
This work has provided a more precise picture of how 2D and 3D object structure is encoded in intermediate and higher-level visual cortices, suggesting that higher- level visual cortex represents categorical identity rather than structure.
Statistically defined visual chunks engage object-based attention
TLDR
The study reports that implicitly learned, statistically defined chunks of abstract visual shapes elicit similar object-based perceptual effects as images of true objects with visual boundaries do, and links the emergence of object representations to implicit statistical learning mechanisms.
Visual statistical learning of temporal structures at different hierarchical levels
TLDR
In summary, in order to process a complex scene, the visual system flexibly uses statistical regularities and the hierarchical structure of the scene.
Perceptual learning and representational learning in humans and animals
  • J. Fiser
  • Psychology, Biology
    Learning & behavior
  • 2009
TLDR
Results of several learning experiments are presented, exploring how internal representations of complex unknown visual patterns might emerge in the brain, and evidence that this learning cannot be captured fully by any simple pairwise associative learning scheme is provided.
Multiple processes in two-dimensional visual statistical learning
TLDR
The results show that subjects are able to learn rules of two-dimensional visual patterns (exemplars) and make categorical judgment of probes based on knowledge of exemplar-based representation, and suggests the coexistence of configural and element-based processing in exempler-based representations.
The automaticity of visual statistical learning.
TLDR
The experiments reported in this article explore the automaticity of VSL in several ways, using both explicit familiarity and implicit response-time measures, and fuel the conclusion that VSL both is and is not automatic.
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 73 REFERENCES
Statistical learning of new visual feature combinations by infants
  • J. Fiser, R. Aslin
  • Biology
    Proceedings of the National Academy of Sciences of the United States of America
  • 2002
TLDR
It is suggested that, similar to lower-level visual representations, infants learn higher-order visual features based on the statistical coherence of elements within the scenes, thereby allowing them to develop an efficient representation for further associative learning.
Unsupervised Statistical Learning of Higher-Order Spatial Structures from Visual Scenes
TLDR
Unsupervised learning of higher-order statistics provides support for Barlow's theory of visual recognition, which posits that detecting “suspicious coincidences” of elements during recognition is a necessary prerequisite for efficient learning of new visual features.
A multi-layer sparse coding network learns contour coding from natural images
Modeling visual recognition from neurobiological constraints
Statistical learning of higher-order temporal structure from visual shape sequences.
  • J. Fiser, R. Aslin
  • Psychology
    Journal of experimental psychology. Learning, memory, and cognition
  • 2002
TLDR
The results suggest that observers automatically extract multiple statistics of temporal events that are suitable for efficient associative learning of new temporal features.
Integration of Local Features into Global Shapes Monkey and Human fMRI Studies
Learning higher-order structures in natural images
TLDR
A hierarchical probabilistic model for learning higher-order statistical regularities in natural images is presented and could provide theoretical insight into the response properties and computational functions of lower level cortical visual areas.
Dynamic binding in a neural network for shape recognition.
TLDR
A neural network is presented that generates a viewpoint-invariant structural description specifying the object's parts and the relations among them and uses synchrony to parse images into their constituent parts.
Fast perceptual learning in visual hyperacuity.
TLDR
This hypothesis is given support by the demonstration that it is possible to synthesize, from a small number of examples of a given task, a simple network that attains the required performance level.
Minimizing Binding Errors Using Learned Conjunctive Features
TLDR
The results provide a more quantitative basis for understanding when spatially invariant conjunctive features can support unambiguous perception in multiobject scenes, and lead to several insights regarding the properties of visual representations optimized for specific recognition tasks.
...
1
2
3
4
5
...