Gunnar Heikkilä

Measuring multimedia quality in mobile networks with an objective parametric model

The perceived quality of a multimedia stream can be estimated with good accuracy with a parametric no-reference opinion model taking codec, total coded bitrate, packet loss rate, and buffering events as input. A rebuffering event in a multimedia stream severely reduces the perceived quality with up to 1.5 MOS units for a QCIF video. High packet loss rates can reduce the quality with as much as 2.5 MOS units. The effect of these two degradation types has been combined in a non-linear way together with information about codec and coded bitrate to form a parametric no-reference objective opinion model. The model shows good performance with RMSE of 0.33 and correlation coefficient of 0.93.

IP-Based Mobile and Fixed Network Audiovisual Media Services

This article provides a tutorial overview of current approaches for monitoring the quality perceived by users of IP-based audiovisual media services. The article addresses both mobile and fixed network services such as mobile TV or Internet Protocol TV (IPTV). It reviews the different quality models that exploit packet- header-, bit stream-, or signal-information for providing audio, video, and audiovisual quality estimates, respectively. It describes how these models can be applied for real-life monitoring, and how they can be adapted to reflect the information available at the given measurement point. An outlook gives insight into emerging trends for near- and mid-term future requirements and solutions.

HTTP adaptive streaming QoE estimation with ITU-T rec. P. 1203: open databases and software

This paper describes an open dataset and software for ITU-T Ree. P.1203. As the first standardized Quality of Experience model for audiovisual HTTP Adaptive Streaming (HAS), it has been extensively trained and validated on over a thousand audiovisual sequences containing HAS-typical effects (such as stalling, coding artifacts, quality switches). Our dataset comprises four of the 30 official subjective databases at a bitstream feature level. The paper also includes subjective results and the model performance. Our software for the standard was made available to the public, too, and it is used for all the analyses presented. Among other previously unpublished details, we show the significant performance improvements of using bitstream-based models over metadata-based ones for video quality analysis, and the robustness of combining classical models with machine-learning-based approaches for estimating user QoE.