Redefining ITU-T P.912 recommendation requirements for subjects of quality assessments in recognition tasks

Abstract

The transmission and analysis of video is often used for a variety of applications outside the entertainment sector, and generally this class of video is used to perform a specific task. Therefore it is crucial to measure, and ultimately, optimize task-based video quality. To develop accurate objective measurements and models for video quality assessment, subjective experiments must be performed. Problems of quality measurements for task-based video are partially addressed in a few preliminary standards and a Recommendation (ITU-T P.912, "Subjective Video Quality Assessment Methods for Recognition Tasks,") that mainly introduce basic definitions, methods of testing and requirements for subjects taking part in psychophysical experiments. Nevertheless, to the best of the authors' knowledge, the issue of requirements for subjects has been not verified in any specific academic research. Consequently, in this paper, we compare groups of subjects assessing video quality for task-based video. Once a comparison has been made for task-based video, specifications amendments for P.912 are developed. These will assist researchers of task-based video quality in identifying the subjects that will allow them to successfully perform the psychophysical experiment required. © 2012 Springer-Verlag.