Discrete quality assessment in IPTV content distribution networks

Maintaining the quality of videos in resource-intensive IPTV services is challenging due to the nature of packet-based content distribution networks (CDN). Network impairments are unpredictable and highly detrimental to the quality of video content. Quality of the end user experience (QoE) has become a critical service differentiator. An efficient real-time quality assessment service in distribution networks is the foundation of service quality monitoring and management. The perceptual impact of individual impairments varies significantly and is influenced by complex impact factors. Without differentiating the impact of quality violation events to the user experience, existing assessment methodologies based on network QoS such as packet loss rate cannot provide adequate supports for the IPTV service assessment. A discrete perceptual impact evaluation quality assessment (DEQA) framework is introduced in this paper. The proposed framework enables a real-time, non-intrusive assessment service by efficiently recognising and assessing individual quality violation events in the IPTV distribution network. The discrete perceptual impacts to a media session are aggregated for the overall user level quality evaluation. With its deployment scheme the DEQA framework also facilitates efficient network diagnosis and QoE management. To realise the key assessment function of the framework and investigate the proposed advanced packet inspection mechanism, we also introduce the dedicated evaluation testbed—the LA2 system. A subjective experiment with data analysis is also presented to demonstrate the development of perceptual impact assessment functions using analytical inference, the tools of the LA2 system, subjective user tests and statistical modelling.     

Quality perception when streaming video on tablet devices

The proposed work aims at analyzing the quality perceived by the user when streaming video on tablet devices. The contributions of this paper are: (i) to analyze the results of subjective quality assessments to determine which Quality of Service (QoS) parameters mainly affect the users’ Quality of Experience (QoE) in video streaming over tablet devices; (ii) to define a parametric quality model useful in system control and optimization for the considered scenarios; (iii) to compare the performance of the proposed model with subjective quality results obtained in alternative state-of-the-art studies and investigate whether other models could be applied to our case and vice versa.     

基于深层特征学习的无参考立体图像质量评价

提出了一种基于深层特征学习的无参考(NR)立体图 像质量评价方 法。与传统人工提取图像特征不同,采用卷积神经网络(CNN)自动提取图像特征,评价过程 分为训练和 测试两阶段。在训练阶段,将图像分块训练CNN网络,利用CNN提取图像块特征,并结合不同 的整合方式 得到图像的全局特征,通过支持向量回归(SVR)建立主观质量与全局特征的回归模型;在测 试阶段,由已训练的CNN网 络和回归模型,得到左右图像和独眼图的质量。最后,根据人眼双目视觉特性融合左图像、 右图像和独眼 图的质量,得到立体图像质量。本文方法在LIVE-I和LIVE-II数据库上的Spearman等级系 数(SROCC)分别达 到了0.94,评价结果准确,与人眼的主 观感受一致。     

基于新型三元卷积神经网络的行人再辨识算法

基于三元卷积神经网络的行人再辨识算法多数采用欧式距离度量行人之间的相似度,并配合铰链(hinge)损失函数进行卷积神经网络的训练。然而,这种作法存在两个不足：欧式距离作为行人相似度,鉴别力不够强;铰链损失函数的间隔(Margin)参数设定依赖于人工预先设定且在训练过程中无法自适应调整。为此,针对上述两个不足进行改进,该文提出一种基于新型三元卷积神经网络的行人再辨识算法,以提高行人再辨识的准确率。首先,提出一种归一化混合度量函数取代传统的度量方法进行行人相似度计算,提高了行人相似度度量的鉴别力;其次,提出采用Log-logistic函数代替铰链函数,无需人工设定间隔参数,改进了特征与度量函数的联合优化效果。实验结果表明,所提出的算法在Auto Detected CUHK03 和VIPeR两个数据库上的准确率均获得显著的提升,验证了所提出算法的优越性。     

Evaluating feedback devices for time-continuous mobile multimedia quality assessment

In January 2014, the new ITU-T P.913 recommendation for measuring subjective video, audio and multimedia quality in any environment has been published. This document does not contain any time-continuous subjective method. However, environmental parameter values are changing continuously in a majority of outdoor and also most indoor environments. To be aware of their impact on the perceived quality, a time-continuous quality assessment methodology is necessary. In previous standards, targeting laboratory-based test settings, a desk-mounted slider of substantial size is recommended. Unfortunately, there are many environments where such a device cannot be used.In this paper, new feedback tools for mobile time-continuous rating are presented and analysed. We developed several alternatives to the generally adopted desk-mounted slider as a rating device. In order to compare the tools, we defined a number of performance measures that can be used in further studies. The suitability and efficacy of the rating scheme based on measurable parameters as well as user opinions is compared. One method, the finger count, seems to outperform the others from all points of view. It was been judged to be easy to use with low potential for distractions. Furthermore, it reaches a similar precision level as the slider, while requiring lower user reaction and scoring times. Low reaction times are particularly important for time-continuous quality assessment, where the reliability of a mapping between impairments and user ratings plays an essential role.     

Evaluation of temporal variation of video quality in packet loss networks

We examine the effect that variations in the temporal quality of videos have on global video quality. We also propose a general framework for constructing temporal video quality assessment (QA) algorithms that seek to assess transient temporal errors, such as packet losses. The proposed framework modifies simple frame-based quality assessment algorithms by incorporating a temporal quality variance factor. We use packet loss from channel errors as a specific study of practical significance. Using the PSNR and the SSIM index as exemplars, we are able to show that the new video QA algorithms are highly responsive to packet loss errors.     

On quality of experience of scalable video adaptation

In this paper, we study the quality of experience (QoE) issues in scalable video coding (SVC) for its adaptation in video communications. A QoE assessment database is developed according to SVC scalabilities. Based on the subjective evaluation results, we derive the optimal scalability adaptation track for the individual video and further summarize common scalability adaptation tracks for videos according to their spatial information (SI) and temporal information (TI). Based on the summarized adaptation tracks, we conclude some general guidelines for the effective SVC video adaptation. A rate-QoE model for SVC adaptation is derived accordingly. Experimental results show that the proposed QoE-aware scalability adaptation scheme significantly outperforms the conventional adaptation schemes in terms of QoE. Moreover, the proposed QoE model reflects the rate and QoE relationship in SVC adaptation and thus, provides a useful methodology to estimate video QoE which is important for QoE-aware scalable video streaming.     

Evaluation of objective quality metrics for multidimensional video scalability

Multidimensional video scalability refers to the possibility that a video sequence can be adapted according to given conditions of video consumption by adjusting one or more of its features such as frame size, frame rate, and spatial quality. An important issue in implementing an adaptive video distribution scheme using scalability is how to maximize the quality of experience for the delivered contents, which raises a more fundamental issue, that is, how to estimate perceived quality of scalable video contents. This paper evaluates existing state-of-the-art objective quality metrics, including both generic image/video metrics and ones particularly developed for scalable videos, on the problem of quality assessment of multidimensional video scalability. It is shown that, on the whole, some recently developed metrics targeting scalability perform best. The results are thoroughly discussed in relation to the nature of the problem in comparison to what has been reported in existing studies for other problems.