With the rapid growth of streaming media applications, there has been a strong demand of Quality-of-Experience (QoE) measurement and QoE-driven video delivery technologies. While the new worldwide standard dynamic adaptive streaming over hypertext transfer protocol (DASH) provides an inter-operable solution to overcome the volatile network conditions, its complex characteristic brings new challenges to the objective video QoE measurement models. How streaming activities such as stalling and bitrate switching events affect QoE is still an open question, and is hardly taken into consideration in the traditionally QoE models. More importantly, with an increasing number of objective QoE models proposed, it is important to evaluate the performance of these algorithms in a comparative setting and analyze the strengths and weaknesses of these methods.In this study, we build two subject-rated streaming video databases. The progressive streaming video database is dedicated to investigate the human responses to the combined effect of video compression, initial buffering, and stalling. The adaptive streaming video database is designed to evaluate the performance of adaptive bitrate streaming algorithms and objective QoE models. We also provide useful insights on the improvement of adaptive bitrate streaming algorithms.Furthermore, we propose a novel QoE prediction approach to account for the instantaneous quality degradation due to perceptual video presentation impairment, the playback stalling events, and the instantaneous interactions between them. Twelve QoE algorithms from four categories including signal fidelity-based, network QoS-based, application QoS-based, and hybrid QoE models are assessed in terms of correlation with human perception on the two streaming video databases. Experimental results show that the proposed model is in close agreement with subjective opinions and significantly outperforms traditional QoE models.
展开▼
