Improved foveation- and saliency-based visual attention prediction under a quality assessment task

Milind S. Gide, Lina Karam

Research output: Chapter in Book/Report/Conference proceedingConference contribution

6 Citations (Scopus)

Abstract

Image quality assessment is one application out of many that can be aided by the use of computational saliency models. Existing visual saliency models have not been extensively tested under a quality assessment context. Also, these models are typically geared towards predicting saliency in non-distorted images. Recent work has also focussed on mimicking the human visual system in order to predict fixation points from saliency maps. One such technique (GAFFE) that uses foveation has been found to perform well for non-distorted images. This work extends the foveation framework by integrating it with saliency maps from well known saliency models. The performance of the foveated saliency models is evaluated based on a comparison with human ground-truth eye-tracking data. For comparison, the performance of the original non-foveated saliency predictions is also presented. It is shown that the integration of saliency models with a foveation based fixation finding framework significantly improves the prediction performance of existing saliency models over different distortion types. It is also found that the information maximization based saliency maps perform the best consistently over different distortion types and levels under this foveation based framework.

Original languageEnglish (US)
Title of host publication2012 4th International Workshop on Quality of Multimedia Experience, QoMEX 2012
Pages200-205
Number of pages6
DOIs
StatePublished - 2012
Event2012 4th International Workshop on Quality of Multimedia Experience, QoMEX 2012 - Melbourne, VIC, Australia
Duration: Jul 5 2012Jul 7 2012

Other

Other2012 4th International Workshop on Quality of Multimedia Experience, QoMEX 2012
CountryAustralia
CityMelbourne, VIC
Period7/5/127/7/12

Fingerprint

Image quality

Keywords

  • Foveation
  • Gaze
  • Information Maximization
  • Quality Assessment
  • Visual Attention

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Computer Vision and Pattern Recognition
  • Software

Cite this

Gide, M. S., & Karam, L. (2012). Improved foveation- and saliency-based visual attention prediction under a quality assessment task. In 2012 4th International Workshop on Quality of Multimedia Experience, QoMEX 2012 (pp. 200-205). [6263871] https://doi.org/10.1109/QoMEX.2012.6263871

Improved foveation- and saliency-based visual attention prediction under a quality assessment task. / Gide, Milind S.; Karam, Lina.

2012 4th International Workshop on Quality of Multimedia Experience, QoMEX 2012. 2012. p. 200-205 6263871.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Gide, MS & Karam, L 2012, Improved foveation- and saliency-based visual attention prediction under a quality assessment task. in 2012 4th International Workshop on Quality of Multimedia Experience, QoMEX 2012., 6263871, pp. 200-205, 2012 4th International Workshop on Quality of Multimedia Experience, QoMEX 2012, Melbourne, VIC, Australia, 7/5/12. https://doi.org/10.1109/QoMEX.2012.6263871
Gide MS, Karam L. Improved foveation- and saliency-based visual attention prediction under a quality assessment task. In 2012 4th International Workshop on Quality of Multimedia Experience, QoMEX 2012. 2012. p. 200-205. 6263871 https://doi.org/10.1109/QoMEX.2012.6263871
Gide, Milind S. ; Karam, Lina. / Improved foveation- and saliency-based visual attention prediction under a quality assessment task. 2012 4th International Workshop on Quality of Multimedia Experience, QoMEX 2012. 2012. pp. 200-205
@inproceedings{1154d8b406dc46cea07fd2475c1164da,
title = "Improved foveation- and saliency-based visual attention prediction under a quality assessment task",
abstract = "Image quality assessment is one application out of many that can be aided by the use of computational saliency models. Existing visual saliency models have not been extensively tested under a quality assessment context. Also, these models are typically geared towards predicting saliency in non-distorted images. Recent work has also focussed on mimicking the human visual system in order to predict fixation points from saliency maps. One such technique (GAFFE) that uses foveation has been found to perform well for non-distorted images. This work extends the foveation framework by integrating it with saliency maps from well known saliency models. The performance of the foveated saliency models is evaluated based on a comparison with human ground-truth eye-tracking data. For comparison, the performance of the original non-foveated saliency predictions is also presented. It is shown that the integration of saliency models with a foveation based fixation finding framework significantly improves the prediction performance of existing saliency models over different distortion types. It is also found that the information maximization based saliency maps perform the best consistently over different distortion types and levels under this foveation based framework.",
keywords = "Foveation, Gaze, Information Maximization, Quality Assessment, Visual Attention",
author = "Gide, {Milind S.} and Lina Karam",
year = "2012",
doi = "10.1109/QoMEX.2012.6263871",
language = "English (US)",
isbn = "9781467307253",
pages = "200--205",
booktitle = "2012 4th International Workshop on Quality of Multimedia Experience, QoMEX 2012",

}

TY - GEN

T1 - Improved foveation- and saliency-based visual attention prediction under a quality assessment task

AU - Gide, Milind S.

AU - Karam, Lina

PY - 2012

Y1 - 2012

N2 - Image quality assessment is one application out of many that can be aided by the use of computational saliency models. Existing visual saliency models have not been extensively tested under a quality assessment context. Also, these models are typically geared towards predicting saliency in non-distorted images. Recent work has also focussed on mimicking the human visual system in order to predict fixation points from saliency maps. One such technique (GAFFE) that uses foveation has been found to perform well for non-distorted images. This work extends the foveation framework by integrating it with saliency maps from well known saliency models. The performance of the foveated saliency models is evaluated based on a comparison with human ground-truth eye-tracking data. For comparison, the performance of the original non-foveated saliency predictions is also presented. It is shown that the integration of saliency models with a foveation based fixation finding framework significantly improves the prediction performance of existing saliency models over different distortion types. It is also found that the information maximization based saliency maps perform the best consistently over different distortion types and levels under this foveation based framework.

AB - Image quality assessment is one application out of many that can be aided by the use of computational saliency models. Existing visual saliency models have not been extensively tested under a quality assessment context. Also, these models are typically geared towards predicting saliency in non-distorted images. Recent work has also focussed on mimicking the human visual system in order to predict fixation points from saliency maps. One such technique (GAFFE) that uses foveation has been found to perform well for non-distorted images. This work extends the foveation framework by integrating it with saliency maps from well known saliency models. The performance of the foveated saliency models is evaluated based on a comparison with human ground-truth eye-tracking data. For comparison, the performance of the original non-foveated saliency predictions is also presented. It is shown that the integration of saliency models with a foveation based fixation finding framework significantly improves the prediction performance of existing saliency models over different distortion types. It is also found that the information maximization based saliency maps perform the best consistently over different distortion types and levels under this foveation based framework.

KW - Foveation

KW - Gaze

KW - Information Maximization

KW - Quality Assessment

KW - Visual Attention

UR - http://www.scopus.com/inward/record.url?scp=84866674659&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84866674659&partnerID=8YFLogxK

U2 - 10.1109/QoMEX.2012.6263871

DO - 10.1109/QoMEX.2012.6263871

M3 - Conference contribution

SN - 9781467307253

SP - 200

EP - 205

BT - 2012 4th International Workshop on Quality of Multimedia Experience, QoMEX 2012

ER -