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

Object detection through search with a foveated visual system.

Emre Akbas1,2, Miguel P Eckstein1,3

  • 1Department of Psychological and Brain Sciences, University of California, Santa Barbara, Santa Barbara, California, United States of America.

Plos Computational Biology
|October 10, 2017
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

Distinct roles of central and peripheral vision in rapid scene understanding.

Journal of vision·2026
Same author

Learned foveal representations, not gaze-mediated information access, underlie perceptual advantages for canonical face configurations.

Journal of vision·2026
Same author

Relating the perceived useful field of view to visual search in 2D Images and 3D volumetric images.

Attention, perception & psychophysics·2026
Same author

Eye movements during free viewing to maximize scene understanding.

Nature communications·2025
Same author

The psychophysics of dynamic gaze-following saccades during search.

Journal of vision·2025
Same author

Emergent neuronal mechanisms mediating covert attention in convolutional neural networks.

Proceedings of the National Academy of Sciences of the United States of America·2025
Same journal

Development of whole-limb skeletal patterning through the coordination of growth and self-organization models.

PLoS computational biology·2026
Same journal

The energetic cost of human standing balance and gait initiation over a range of natural postures.

PLoS computational biology·2026
Same journal

Combinatorial multiomic analysis from a pedigree of Sox10Dom Hirschsprung mice identifies multiple high confidence candidate modifiers of Enteric Nervous System development.

PLoS computational biology·2026
Same journal

Extracting host-specific developmental signatures from longitudinal microbiome data.

PLoS computational biology·2026
Same journal

Population sparseness determines strength of Hebbian plasticity for maximal memory lifetime in associative networks.

PLoS computational biology·2026
Same journal

Predictive coding explains asymmetric connectivity in the brain: A neural network study.

PLoS computational biology·2026
See all related articles

Foveated vision, with its varying spatial resolution, maintains object detection performance while significantly reducing computational costs. This strategy, mimicking human visual systems, offers metabolic savings without compromising visual search accuracy.

Area of Science:

  • Computational Neuroscience
  • Computer Vision
  • Visual Perception

Background:

  • Biological visual systems exhibit foveated vision, processing scenes with varying spatial resolution and utilizing eye movements for efficient information gathering.
  • The computational and metabolic advantages of foveated vision are well-established, but its performance costs in tasks like visual search require further investigation.

Purpose of the Study:

  • To develop and evaluate a foveated object detector that mimics biological vision by processing scenes with non-uniform spatial resolution.
  • To compare the performance and computational efficiency of the foveated object detector against a non-foveated, high-resolution baseline.
  • To assess the impact of foveation on bottom-up saliency computation and its agreement with non-foveated models.

Main Methods:

Related Experiment Videos

  • Developed a foveated object detector integrating computer vision object detectors with a model of peripheral pooling regions in the human visual system's V1 layer.
  • Implemented retino-specific classifiers to guide eye movements, aligning the fovea with regions of interest and integrating information across fixations.
  • Compared the foveated detector's accuracy in identifying 20 object classes against a non-foveated detector using the PASCAL VOC 2007 dataset.

Main Results:

  • The foveated object detector achieved performance comparable to the non-foveated, high-resolution detector.
  • Significant computational cost savings were observed with the foveated approach.
  • A foveated bottom-up saliency model demonstrated agreement with non-foveated models in identifying salient regions, suggesting preserved perceptual performance.

Conclusions:

  • Foveated vision, combined with eye movements, can preserve perceptual performance in visual search tasks while offering substantial computational and metabolic savings.
  • These findings support the evolutionary advantage of foveated visual systems as an efficient strategy for processing complex visual information.
  • The developed foveated object detector serves as a valuable tool for understanding the trade-offs between computational cost and performance in visual processing.