Pause intensity: a no-reference quality assessment metric for video streaming in TCP networks

Colin Bailey*, Mirghiasaldin Seyedebrahimi, Xiao-Hong Peng

*Corresponding author for this work

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

Abstract

In this paper a full analytic model for pause intensity (PI), a no-reference metric for video quality assessment, is presented. The model is built upon the video play out buffer behavior at the client side and also encompasses the characteristics of a TCP network. Video streaming via TCP produces impairments in play continuity, which are not typically reflected in current objective metrics such as PSNR and SSIM. Recently the buffer under run frequency/probability has been used to characterize the buffer behavior and as a measurement for performance optimization. But we show, using subjective testing, that under run frequency cannot reflect the viewers' quality of experience for TCP based streaming. We also demonstrate that PI is a comprehensive metric made up of a combination of phenomena observed in the play out buffer. The analytical model in this work is verified with simulations carried out on ns-2, showing that the two results are closely matched. The effectiveness of the PI metric has also been proved by subjective testing on a range of video clips, where PI values exhibit a good correlation with the viewers' opinion scores.

Original languageEnglish
Title of host publicationProceedings, 2012 IEEE International Conference on Multimedia and Expo
Subtitle of host publicationICME 2012
PublisherIEEE
Pages818-823
Number of pages6
ISBN (Electronic)978-1-4673-1659-0
ISBN (Print)978-0-7695-4711-4
DOIs
Publication statusPublished - 2012
Event2012 IEEE International Conference on Multimedia and Expo - Melbourne, Australia
Duration: 9 Jul 201213 Jul 2012

Publication series

NameProceedings
PublisherIEEE
ISSN (Print)1945-7871
ISSN (Electronic)1945-788X

Conference

Conference2012 IEEE International Conference on Multimedia and Expo
Abbreviated titleICME 2012
CountryAustralia
CityMelbourne
Period9/07/1213/07/12

Fingerprint

Subjective testing
Video streaming
Analytical models

Keywords

  • assessment metric
  • pause intensity
  • TCP networks
  • video quality
  • video transmission
  • keywords-component

Cite this

Bailey, C., Seyedebrahimi, M., & Peng, X-H. (2012). Pause intensity: a no-reference quality assessment metric for video streaming in TCP networks. In Proceedings, 2012 IEEE International Conference on Multimedia and Expo: ICME 2012 (pp. 818-823). (Proceedings). IEEE. https://doi.org/10.1109/ICME.2012.148
Bailey, Colin ; Seyedebrahimi, Mirghiasaldin ; Peng, Xiao-Hong. / Pause intensity : a no-reference quality assessment metric for video streaming in TCP networks. Proceedings, 2012 IEEE International Conference on Multimedia and Expo: ICME 2012. IEEE, 2012. pp. 818-823 (Proceedings).
@inproceedings{dac903b5ab654e4eb3af722939f009dc,
title = "Pause intensity: a no-reference quality assessment metric for video streaming in TCP networks",
abstract = "In this paper a full analytic model for pause intensity (PI), a no-reference metric for video quality assessment, is presented. The model is built upon the video play out buffer behavior at the client side and also encompasses the characteristics of a TCP network. Video streaming via TCP produces impairments in play continuity, which are not typically reflected in current objective metrics such as PSNR and SSIM. Recently the buffer under run frequency/probability has been used to characterize the buffer behavior and as a measurement for performance optimization. But we show, using subjective testing, that under run frequency cannot reflect the viewers' quality of experience for TCP based streaming. We also demonstrate that PI is a comprehensive metric made up of a combination of phenomena observed in the play out buffer. The analytical model in this work is verified with simulations carried out on ns-2, showing that the two results are closely matched. The effectiveness of the PI metric has also been proved by subjective testing on a range of video clips, where PI values exhibit a good correlation with the viewers' opinion scores.",
keywords = "assessment metric, pause intensity, TCP networks, video quality, video transmission, keywords-component",
author = "Colin Bailey and Mirghiasaldin Seyedebrahimi and Xiao-Hong Peng",
year = "2012",
doi = "10.1109/ICME.2012.148",
language = "English",
isbn = "978-0-7695-4711-4",
series = "Proceedings",
publisher = "IEEE",
pages = "818--823",
booktitle = "Proceedings, 2012 IEEE International Conference on Multimedia and Expo",
address = "United States",

}

Bailey, C, Seyedebrahimi, M & Peng, X-H 2012, Pause intensity: a no-reference quality assessment metric for video streaming in TCP networks. in Proceedings, 2012 IEEE International Conference on Multimedia and Expo: ICME 2012. Proceedings, IEEE, pp. 818-823, 2012 IEEE International Conference on Multimedia and Expo, Melbourne, Australia, 9/07/12. https://doi.org/10.1109/ICME.2012.148

Pause intensity : a no-reference quality assessment metric for video streaming in TCP networks. / Bailey, Colin; Seyedebrahimi, Mirghiasaldin; Peng, Xiao-Hong.

Proceedings, 2012 IEEE International Conference on Multimedia and Expo: ICME 2012. IEEE, 2012. p. 818-823 (Proceedings).

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

TY - GEN

T1 - Pause intensity

T2 - a no-reference quality assessment metric for video streaming in TCP networks

AU - Bailey, Colin

AU - Seyedebrahimi, Mirghiasaldin

AU - Peng, Xiao-Hong

PY - 2012

Y1 - 2012

N2 - In this paper a full analytic model for pause intensity (PI), a no-reference metric for video quality assessment, is presented. The model is built upon the video play out buffer behavior at the client side and also encompasses the characteristics of a TCP network. Video streaming via TCP produces impairments in play continuity, which are not typically reflected in current objective metrics such as PSNR and SSIM. Recently the buffer under run frequency/probability has been used to characterize the buffer behavior and as a measurement for performance optimization. But we show, using subjective testing, that under run frequency cannot reflect the viewers' quality of experience for TCP based streaming. We also demonstrate that PI is a comprehensive metric made up of a combination of phenomena observed in the play out buffer. The analytical model in this work is verified with simulations carried out on ns-2, showing that the two results are closely matched. The effectiveness of the PI metric has also been proved by subjective testing on a range of video clips, where PI values exhibit a good correlation with the viewers' opinion scores.

AB - In this paper a full analytic model for pause intensity (PI), a no-reference metric for video quality assessment, is presented. The model is built upon the video play out buffer behavior at the client side and also encompasses the characteristics of a TCP network. Video streaming via TCP produces impairments in play continuity, which are not typically reflected in current objective metrics such as PSNR and SSIM. Recently the buffer under run frequency/probability has been used to characterize the buffer behavior and as a measurement for performance optimization. But we show, using subjective testing, that under run frequency cannot reflect the viewers' quality of experience for TCP based streaming. We also demonstrate that PI is a comprehensive metric made up of a combination of phenomena observed in the play out buffer. The analytical model in this work is verified with simulations carried out on ns-2, showing that the two results are closely matched. The effectiveness of the PI metric has also been proved by subjective testing on a range of video clips, where PI values exhibit a good correlation with the viewers' opinion scores.

KW - assessment metric

KW - pause intensity

KW - TCP networks

KW - video quality

KW - video transmission

KW - keywords-component

UR - http://www.scopus.com/inward/record.url?scp=84868128307&partnerID=8YFLogxK

UR - http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6298504

U2 - 10.1109/ICME.2012.148

DO - 10.1109/ICME.2012.148

M3 - Conference contribution

AN - SCOPUS:84868128307

SN - 978-0-7695-4711-4

T3 - Proceedings

SP - 818

EP - 823

BT - Proceedings, 2012 IEEE International Conference on Multimedia and Expo

PB - IEEE

ER -

Bailey C, Seyedebrahimi M, Peng X-H. Pause intensity: a no-reference quality assessment metric for video streaming in TCP networks. In Proceedings, 2012 IEEE International Conference on Multimedia and Expo: ICME 2012. IEEE. 2012. p. 818-823. (Proceedings). https://doi.org/10.1109/ICME.2012.148