Objective Assessment of Region of Interest-Aware Adaptive Multimedia Streaming Quality

Adaptive multimedia streaming relies on controlled adjustment of content bitrate and consequent video quality variation in order to meet the bandwidth constraints of the communication link used for content delivery to the end-user. The values of the easy to measure network-related Quality of Service metrics have no direct relationship with the way moving images are perceived by the human viewer. Consequently variations in the video stream bitrate are not clearly linked to similar variation in the user perceived quality. This is especially true if some human visual system-based adaptation techniques are employed. As research has shown, there are certain image regions in each frame of a video sequence on which the users are more interested than in the others. This paper presents the Region of Interest-based Adaptive Scheme (ROIAS) which adjusts differently the regions within each frame of the streamed multimedia content based on the user interest in them. ROIAS is presented and discussed in terms of the adjustment algorithms employed and their impact on the human perceived video quality. Comparisons with existing approaches, including a constant quality adaptation scheme across the whole frame area, are performed employing two objective metrics which estimate user perceived video quality.