A Caltech Library Service

On the Optimality of Spatial Attention for Object Detection

Harel, Jonathan and Koch, Christof (2009) On the Optimality of Spatial Attention for Object Detection. In: Attention in Cognitive Systems. Lecture Notes in Artificial Intelligence . No.5395. Springer , pp. 1-14. ISBN 978-3-642-00581-7.

Full text is not posted in this repository. Consult Related URLs below.

Use this Persistent URL to link to this item:


Studies on visual attention traditionally focus on its physiological and psychophysical nature [16,18,19], or its algorithmic applications [1,9,21]. We here develop a simple, formal mathematical model of the advantage of spatial attention for object detection, in which spatial attention is defined as processing a subset of the visual input, and detection is an abstraction with certain failure characteristics. We demonstrate that it is suboptimal to process the entire visual input given prior information about target locations, which in practice is almost always available in a video setting due to tracking, motion, or saliency. This argues for an attentional strategy independent of computational savings: no matter how much computational power is available, it is in principle better to dedicate it preferentially to selected portions of the scene. This suggests, anecdotally, a form of environmental pressure for the evolution of foveated photoreceptor densities in the retina. It also offers a general justification for the use of spatial attention in machine vision.

Item Type:Book Section
Related URLs:
URLURL TypeDescription ReadCube access
Koch, Christof0000-0001-6482-8067
Additional Information:© 2009 Springer-Verlag Berlin Heidelberg. We wish to thank DARPA for its generous support of a research program for the development of a biologically modeled object recognition system, and our close collaborators on that program, Sharat Chikkerur at MIT, and Rob Peters at USC.
Group:Koch Laboratory (KLAB)
Funding AgencyGrant Number
Defense Advanced Research Projects Agency (DARPA)UNSPECIFIED
Series Name:Lecture Notes in Artificial Intelligence
Issue or Number:5395
Record Number:CaltechAUTHORS:20100625-153044964
Persistent URL:
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:18819
Deposited By: Tony Diaz
Deposited On:02 Aug 2010 21:54
Last Modified:05 May 2020 18:52

Repository Staff Only: item control page