Abstract

The human visual system (HVS) has evolved to have the ability to selectively focus on the most relevant parts of a visual scene. This mechanism, referred to as visual attention (VA), has been the focus of several neurological and psychological studies in the past few decades. These studies have inspired several computational VA models which have been successfully applied to problems in computer vision and robotics. In this paper we provide a comprehensive survey of the state-of-the-art in computational VA modeling with a special focus on the latest trends. We review several models published since 2012. We also discuss theoretical advantages and disadvantages of each approach. In addition, we describe existing methodologies to evaluate computational models through the use of eye-tracking data along with the VA performance metrics used. We also discuss shortcomings in existing approaches and describe approaches to overcome these shortcomings. A recent subjective evaluation for benchmarking existing VA metrics is also presented and open problems in VA are discussed.

Original languageEnglish (US)
Pages (from-to)347-427
Number of pages81
JournalFoundations and Trends in Signal Processing
Volume10
Issue number4
DOIs
StatePublished - 2017

Fingerprint

Benchmarking
Computer vision
Robotics

ASJC Scopus subject areas

  • Signal Processing

Cite this

Computational visual attention models. / Gide, Milind S.; Karam, Lina.

In: Foundations and Trends in Signal Processing, Vol. 10, No. 4, 2017, p. 347-427.

Research output: Contribution to journalArticle

@article{e877815ff2ae4a82af0c36cee7e32ede,
title = "Computational visual attention models",
abstract = "The human visual system (HVS) has evolved to have the ability to selectively focus on the most relevant parts of a visual scene. This mechanism, referred to as visual attention (VA), has been the focus of several neurological and psychological studies in the past few decades. These studies have inspired several computational VA models which have been successfully applied to problems in computer vision and robotics. In this paper we provide a comprehensive survey of the state-of-the-art in computational VA modeling with a special focus on the latest trends. We review several models published since 2012. We also discuss theoretical advantages and disadvantages of each approach. In addition, we describe existing methodologies to evaluate computational models through the use of eye-tracking data along with the VA performance metrics used. We also discuss shortcomings in existing approaches and describe approaches to overcome these shortcomings. A recent subjective evaluation for benchmarking existing VA metrics is also presented and open problems in VA are discussed.",
author = "Gide, {Milind S.} and Lina Karam",
year = "2017",
doi = "10.1561/2000000055",
language = "English (US)",
volume = "10",
pages = "347--427",
journal = "Foundations and Trends in Signal Processing",
issn = "1932-8346",
publisher = "Now Publishers Inc",
number = "4",

}

TY - JOUR

T1 - Computational visual attention models

AU - Gide, Milind S.

AU - Karam, Lina

PY - 2017

Y1 - 2017

N2 - The human visual system (HVS) has evolved to have the ability to selectively focus on the most relevant parts of a visual scene. This mechanism, referred to as visual attention (VA), has been the focus of several neurological and psychological studies in the past few decades. These studies have inspired several computational VA models which have been successfully applied to problems in computer vision and robotics. In this paper we provide a comprehensive survey of the state-of-the-art in computational VA modeling with a special focus on the latest trends. We review several models published since 2012. We also discuss theoretical advantages and disadvantages of each approach. In addition, we describe existing methodologies to evaluate computational models through the use of eye-tracking data along with the VA performance metrics used. We also discuss shortcomings in existing approaches and describe approaches to overcome these shortcomings. A recent subjective evaluation for benchmarking existing VA metrics is also presented and open problems in VA are discussed.

AB - The human visual system (HVS) has evolved to have the ability to selectively focus on the most relevant parts of a visual scene. This mechanism, referred to as visual attention (VA), has been the focus of several neurological and psychological studies in the past few decades. These studies have inspired several computational VA models which have been successfully applied to problems in computer vision and robotics. In this paper we provide a comprehensive survey of the state-of-the-art in computational VA modeling with a special focus on the latest trends. We review several models published since 2012. We also discuss theoretical advantages and disadvantages of each approach. In addition, we describe existing methodologies to evaluate computational models through the use of eye-tracking data along with the VA performance metrics used. We also discuss shortcomings in existing approaches and describe approaches to overcome these shortcomings. A recent subjective evaluation for benchmarking existing VA metrics is also presented and open problems in VA are discussed.

UR - http://www.scopus.com/inward/record.url?scp=85029752372&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85029752372&partnerID=8YFLogxK

U2 - 10.1561/2000000055

DO - 10.1561/2000000055

M3 - Article

AN - SCOPUS:85029752372

VL - 10

SP - 347

EP - 427

JO - Foundations and Trends in Signal Processing

JF - Foundations and Trends in Signal Processing

SN - 1932-8346

IS - 4

ER -