Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Similarity-based models of human visual recognition

A Unzicker1, M Jüttner, I Rentschler

  • 1Institute of Medical Psychology, University of Munich, Germany. sascha@imp.med.uni-muenchen.de

Vision Research
|November 3, 1998
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

[The wettability of leaf surfaces and the submicroscopic structure of their wax].

Planta·2014
Same author

[Electron-microscopical investigation on wax-covered stomatas].

Planta·2014
Same author

Amplitude and phase characteristics of the steady-state visual evoked potential.

Applied optics·2010
Same author

Object recognition and image understanding: theories of everything?

Spatial vision·2001
Same author

Object and scene analysis by saccadic eye-movements: an investigation with higher-order statistics.

Spatial vision·2001
Same author

Dynamics and context dependence of visual category learning.

Spatial vision·2001

Comparing seven human visual recognition models, this study found that classification theories need both physical signal and cognitive bias. Virtual prototype models excel in extrafoveal viewing, while hyperBF models offer computational efficiency.

Area of Science:

  • Cognitive Psychology
  • Visual Psychophysics
  • Computational Neuroscience
  • Machine Learning

Background:

  • Human visual recognition involves complex processes influenced by both physical stimuli and internal cognitive factors.
  • Existing models of visual recognition stem from diverse fields, including cognitive psychology, visual psychophysics, and connectionism.
  • Evaluating these models requires rigorous comparison against empirical psychophysical data.

Purpose of the Study:

  • To compare the predictive performance of seven distinct human visual recognition models.
  • To assess model accuracy under varying viewing conditions (foveal vs. extrafoveal) and stimulus sets.
  • To identify which models best account for stimulus- and observer-dependencies in visual classification.

Main Methods:

Related Experiment Videos

  • Supervised learning was employed to train models using parametrised grey-level patterns (compound Gabor signals).
  • Model performance was evaluated by comparing predicted data against observed psychophysical classification data.
  • Metrics included root mean square deviation and signal reconstruction accuracy via multidimensional scaling.

Main Results:

  • A successful psychophysical classification theory necessitates a similarity concept integrating physical signal properties and cognitive bias.
  • Model performance was comparable in foveal recognition but diverged significantly in extrafoveal recognition, highlighting the role of cognitive bias.
  • Virtual prototype models demonstrated superior accommodation of stimulus- and observer-dependencies, particularly in extrafoveal viewing.

Conclusions:

  • Cognitive bias plays a crucial role in visual recognition, especially under peripheral (extrafoveal) viewing conditions.
  • Virtual prototype models offer an advantageous framework for understanding human visual classification due to their ability to incorporate observer-specific factors.
  • Computational efficiency varied across models, with hyperBF models being notably fast and generalized signal detection models being slower.