No Reference Video-Quality-Assessment Model for Monitoring Video Quality of IPTV Services

Kazuhisa YAMAGISHI; Jun OKAMOTO; Takanori HAYASHI; Akira TAKAHASHI

doi:10.1587/transcom.E95.B.435

No Reference Video-Quality-Assessment Model for Monitoring Video Quality of IPTV Services

Kazuhisa YAMAGISHI, Jun OKAMOTO, Takanori HAYASHI, Akira TAKAHASHI

Full Text Views

0

Share
Cite this

Summary :

Service providers should monitor the quality of experience of a communication service in real time to confirm its status. To do this, we previously proposed a packet-layer model that can be used for monitoring the average video quality of typical Internet protocol television content using parameters derived from transmitted packet headers. However, it is difficult to monitor the video quality per user using the average video quality because video quality depends on the video content. To accurately monitor the video quality per user, a model that can be used for estimating the video quality per video content rather than the average video quality should be developed. Therefore, to take into account the impact of video content on video quality, we propose a model that calculates the difference in video quality between the video quality of the estimation-target video and the average video quality estimated using a packet-layer model. We first conducted extensive subjective quality assessments for different codecs and video sequences. We then model their characteristics based on parameters related to compression and packet loss. Finally, we verify the performance of the proposed model by applying it to unknown data sets different from the training data sets used for developing the model.

Publication: IEICE TRANSACTIONS on Communications Vol.E95-B No.2 pp.435-448

Publication Date: 2012/02/01

Publicized

Online ISSN: 1745-1345

DOI: 10.1587/transcom.E95.B.435

Type of Manuscript: PAPER

Category: Network

Cite this

Copy

Kazuhisa YAMAGISHI, Jun OKAMOTO, Takanori HAYASHI, Akira TAKAHASHI, "No Reference Video-Quality-Assessment Model for Monitoring Video Quality of IPTV Services" in IEICE TRANSACTIONS on Communications, vol. E95-B, no. 2, pp. 435-448, February 2012, doi: 10.1587/transcom.E95.B.435.
Abstract: Service providers should monitor the quality of experience of a communication service in real time to confirm its status. To do this, we previously proposed a packet-layer model that can be used for monitoring the average video quality of typical Internet protocol television content using parameters derived from transmitted packet headers. However, it is difficult to monitor the video quality per user using the average video quality because video quality depends on the video content. To accurately monitor the video quality per user, a model that can be used for estimating the video quality per video content rather than the average video quality should be developed. Therefore, to take into account the impact of video content on video quality, we propose a model that calculates the difference in video quality between the video quality of the estimation-target video and the average video quality estimated using a packet-layer model. We first conducted extensive subjective quality assessments for different codecs and video sequences. We then model their characteristics based on parameters related to compression and packet loss. Finally, we verify the performance of the proposed model by applying it to unknown data sets different from the training data sets used for developing the model.
URL: https://globals.ieice.org/en_transactions/communications/10.1587/transcom.E95.B.435/_p

Copy

@ARTICLE{e95-b_2_435,
author={Kazuhisa YAMAGISHI, Jun OKAMOTO, Takanori HAYASHI, Akira TAKAHASHI, },
journal={IEICE TRANSACTIONS on Communications},
title={No Reference Video-Quality-Assessment Model for Monitoring Video Quality of IPTV Services},
year={2012},
volume={E95-B},
number={2},
pages={435-448},
abstract={Service providers should monitor the quality of experience of a communication service in real time to confirm its status. To do this, we previously proposed a packet-layer model that can be used for monitoring the average video quality of typical Internet protocol television content using parameters derived from transmitted packet headers. However, it is difficult to monitor the video quality per user using the average video quality because video quality depends on the video content. To accurately monitor the video quality per user, a model that can be used for estimating the video quality per video content rather than the average video quality should be developed. Therefore, to take into account the impact of video content on video quality, we propose a model that calculates the difference in video quality between the video quality of the estimation-target video and the average video quality estimated using a packet-layer model. We first conducted extensive subjective quality assessments for different codecs and video sequences. We then model their characteristics based on parameters related to compression and packet loss. Finally, we verify the performance of the proposed model by applying it to unknown data sets different from the training data sets used for developing the model.},
keywords={},
doi={10.1587/transcom.E95.B.435},
ISSN={1745-1345},
month={February},}

Copy

TY - JOUR
TI - No Reference Video-Quality-Assessment Model for Monitoring Video Quality of IPTV Services
T2 - IEICE TRANSACTIONS on Communications
SP - 435
EP - 448
AU - Kazuhisa YAMAGISHI
AU - Jun OKAMOTO
AU - Takanori HAYASHI
AU - Akira TAKAHASHI
PY - 2012
DO - 10.1587/transcom.E95.B.435
JO - IEICE TRANSACTIONS on Communications
SN - 1745-1345
VL - E95-B
IS - 2
JA - IEICE TRANSACTIONS on Communications
Y1 - February 2012
AB - Service providers should monitor the quality of experience of a communication service in real time to confirm its status. To do this, we previously proposed a packet-layer model that can be used for monitoring the average video quality of typical Internet protocol television content using parameters derived from transmitted packet headers. However, it is difficult to monitor the video quality per user using the average video quality because video quality depends on the video content. To accurately monitor the video quality per user, a model that can be used for estimating the video quality per video content rather than the average video quality should be developed. Therefore, to take into account the impact of video content on video quality, we propose a model that calculates the difference in video quality between the video quality of the estimation-target video and the average video quality estimated using a packet-layer model. We first conducted extensive subjective quality assessments for different codecs and video sequences. We then model their characteristics based on parameters related to compression and packet loss. Finally, we verify the performance of the proposed model by applying it to unknown data sets different from the training data sets used for developing the model.
ER -