Embedded quality assessment metrics for video presentations

In order to monitor and assess the quality of multimedia over IP we will develop several metrics to measure video quality. In particular no-reference metrics that estimate video quality in the absence of a reference video stream could be very useful in monitoring the quality of streamed video. We need to embed these metrics in the video player in order to improve the perceived quality of video presentations in a Universal Multimedia Access (UMA) enabled system. This leads to our view of QoS driven video players (viewers). At system level we propose to embed these metrics in the delivered bit stream over an IP network so as to use it to enhance the decoded sequence by optimizing deblocking filters and error concealment for error resilient systems. Our investigations of the embedded metrics will form the basis for setting the requirements of novel multimedia viewers.

Relations to external projects and activities:
Coordinated with Research Council project 140445/13 "Universal Multimedia Access from Wired and Wireless Systems" and COST 276 "Information and Knowledge Management for Integrated Media Communications Systems"

Project coordinator: Andrew Perkis
Professor: Andrew Perkis
Postdocs: Ajit S. Bopardikar, Venkatesh Babu Radhakrishnan
PhD student: Odd Inge Hillestad