Performance-analysis-based acceleration of image quality assessment

Thien Phan, Sohum Sohoni, Damon M. Chandler, Eric C. Larson

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

7 Citations (Scopus)

Abstract

Two stages are commonly employed in modern algorithms of image/video quality assessment (QA): (1) a local frequency-based decomposition, and (2) block-based statistical comparisons between the frequency coefficients of the reference and distorted images. This paper presents a performance analysis of and techniques for accelerating these stages. We specifically analyze and accelerate one representative QA algorithm recently developed by the authors (Larson and Chandler, 2010). We identify the bottlenecks from the abovementioned stages, and we present methods of acceleration using integral images, inline expansion, a GPGPU implementation, and other code modifications. We show how a combination of these approaches can yield a speedup of 47x.

Original languageEnglish (US)
Title of host publicationProceedings of the IEEE Southwest Symposium on Image Analysis and Interpretation
Pages81-84
Number of pages4
DOIs
StatePublished - 2012
Externally publishedYes
Event2012 IEEE Southwest Symposium on Image Analysis and Interpretation, SSIAI 2012 - Santa Fe, NM, United States
Duration: Apr 22 2012Apr 24 2012

Other

Other2012 IEEE Southwest Symposium on Image Analysis and Interpretation, SSIAI 2012
CountryUnited States
CitySanta Fe, NM
Period4/22/124/24/12

Fingerprint

Image quality
Decomposition

Keywords

  • acceleration
  • code optimization
  • GPGPU
  • Image quality
  • integral image
  • video quality

ASJC Scopus subject areas

  • Software
  • Computer Vision and Pattern Recognition
  • Computer Science Applications

Cite this

Phan, T., Sohoni, S., Chandler, D. M., & Larson, E. C. (2012). Performance-analysis-based acceleration of image quality assessment. In Proceedings of the IEEE Southwest Symposium on Image Analysis and Interpretation (pp. 81-84). [6202458] https://doi.org/10.1109/SSIAI.2012.6202458

Performance-analysis-based acceleration of image quality assessment. / Phan, Thien; Sohoni, Sohum; Chandler, Damon M.; Larson, Eric C.

Proceedings of the IEEE Southwest Symposium on Image Analysis and Interpretation. 2012. p. 81-84 6202458.

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

Phan, T, Sohoni, S, Chandler, DM & Larson, EC 2012, Performance-analysis-based acceleration of image quality assessment. in Proceedings of the IEEE Southwest Symposium on Image Analysis and Interpretation., 6202458, pp. 81-84, 2012 IEEE Southwest Symposium on Image Analysis and Interpretation, SSIAI 2012, Santa Fe, NM, United States, 4/22/12. https://doi.org/10.1109/SSIAI.2012.6202458
Phan T, Sohoni S, Chandler DM, Larson EC. Performance-analysis-based acceleration of image quality assessment. In Proceedings of the IEEE Southwest Symposium on Image Analysis and Interpretation. 2012. p. 81-84. 6202458 https://doi.org/10.1109/SSIAI.2012.6202458
Phan, Thien ; Sohoni, Sohum ; Chandler, Damon M. ; Larson, Eric C. / Performance-analysis-based acceleration of image quality assessment. Proceedings of the IEEE Southwest Symposium on Image Analysis and Interpretation. 2012. pp. 81-84
@inproceedings{b0bb1df372ab4c16966ab2ae9924ce58,
title = "Performance-analysis-based acceleration of image quality assessment",
abstract = "Two stages are commonly employed in modern algorithms of image/video quality assessment (QA): (1) a local frequency-based decomposition, and (2) block-based statistical comparisons between the frequency coefficients of the reference and distorted images. This paper presents a performance analysis of and techniques for accelerating these stages. We specifically analyze and accelerate one representative QA algorithm recently developed by the authors (Larson and Chandler, 2010). We identify the bottlenecks from the abovementioned stages, and we present methods of acceleration using integral images, inline expansion, a GPGPU implementation, and other code modifications. We show how a combination of these approaches can yield a speedup of 47x.",
keywords = "acceleration, code optimization, GPGPU, Image quality, integral image, video quality",
author = "Thien Phan and Sohum Sohoni and Chandler, {Damon M.} and Larson, {Eric C.}",
year = "2012",
doi = "10.1109/SSIAI.2012.6202458",
language = "English (US)",
isbn = "9781467318303",
pages = "81--84",
booktitle = "Proceedings of the IEEE Southwest Symposium on Image Analysis and Interpretation",

}

TY - GEN

T1 - Performance-analysis-based acceleration of image quality assessment

AU - Phan, Thien

AU - Sohoni, Sohum

AU - Chandler, Damon M.

AU - Larson, Eric C.

PY - 2012

Y1 - 2012

N2 - Two stages are commonly employed in modern algorithms of image/video quality assessment (QA): (1) a local frequency-based decomposition, and (2) block-based statistical comparisons between the frequency coefficients of the reference and distorted images. This paper presents a performance analysis of and techniques for accelerating these stages. We specifically analyze and accelerate one representative QA algorithm recently developed by the authors (Larson and Chandler, 2010). We identify the bottlenecks from the abovementioned stages, and we present methods of acceleration using integral images, inline expansion, a GPGPU implementation, and other code modifications. We show how a combination of these approaches can yield a speedup of 47x.

AB - Two stages are commonly employed in modern algorithms of image/video quality assessment (QA): (1) a local frequency-based decomposition, and (2) block-based statistical comparisons between the frequency coefficients of the reference and distorted images. This paper presents a performance analysis of and techniques for accelerating these stages. We specifically analyze and accelerate one representative QA algorithm recently developed by the authors (Larson and Chandler, 2010). We identify the bottlenecks from the abovementioned stages, and we present methods of acceleration using integral images, inline expansion, a GPGPU implementation, and other code modifications. We show how a combination of these approaches can yield a speedup of 47x.

KW - acceleration

KW - code optimization

KW - GPGPU

KW - Image quality

KW - integral image

KW - video quality

UR - http://www.scopus.com/inward/record.url?scp=84862905227&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84862905227&partnerID=8YFLogxK

U2 - 10.1109/SSIAI.2012.6202458

DO - 10.1109/SSIAI.2012.6202458

M3 - Conference contribution

SN - 9781467318303

SP - 81

EP - 84

BT - Proceedings of the IEEE Southwest Symposium on Image Analysis and Interpretation

ER -