Exploring the acceptability of the audiovisual quality for a mobile video session based on objectively measured parameters

Abstract

Although Quality of Experience (QoE) plays a major role in the design and development of mobile applications and services, QoE assessment is still challenging, especially in real-life (so called `living lab') contexts. This paper presents results from an exploratory, interdisciplinary study on QoE and more specifically, on the acceptability of varying audiovisual qualities during video watching on a mobile device, in a realistic setting. The subjectively measured acceptability is modeled by means of a decision tree which uses information about the network type, transport protocol, technical video characteristics, and user behavior. The proposed model may contribute to the estimation of users' subjective evaluation of the quality during mobile video watching and to QoE optimization by dynamically altering the parameters that have the largest influence on this subjective evaluation

Acceptability-Based QoE Models for Mobile Video

“…These models do not consider other influencing aspects, such as video resolution, and are dependent on the ability of knowing the video content types. In [12], a decision tree of audiovisual quality acceptance is determined by means of network type, transport protocol, video quality, and user watching behavior. The research made in living context, yet the subjective assessments for only two video qualities may not be enough to delegate the usage situations.…”

“…The mapping process is to find the best-fit quality prediction model. There are many different approaches, including discriminant analysis [14], machine learning classification algorithm -decision tree [11,12], multiple linear regression [15], and non-linear regression [4,6]. Discriminant analysis and decision tree are suitable for classifying groups, such as acceptable or unacceptable, whereas linear and non-linear regression are appropriate for calculating an index, such as acceptability and MOS.…”

“…In general, taking account the mobile screen feature into the QoE models provides more accurate prediction, which can be seen by comparing the performance of (11), (13) and (15) with (12), (14) and (16), respectively. The important influence of device screen resolution on prediction accuracy can also be visualized through Fig.…”

“…Telecommunication standardization sector of International Telecommunication Union (ITU-T) defines QoE as the end-user's overall acceptability of a service or application [9]. However, a little research has focused on establishing models to predict the user acceptance of mobile video [7,[10][11][12]. The work could be improved by involving more influencing factors of user experience such as device characteristics and use context.…”

Impact of mobile devices and usage location on perceived multimedia quality

“…They varied codec rates to create eight quality levels and found that mobile device users will deem multimedia quality acceptable differently on different devices. Research has been done [10] using mobile devices in "living labs"-letting experiment participants use a mobile device during their daily routine. Delivery of multimedia content to a mobile phone using Digital Video Broadcasting for Handhelds (DVB-H) was tested in the context of a train station, a bus, and a café in [11].…”