Video streaming applications have become immensely popular, leading to increasing user expectations for high-quality services. Extensive work has been conducted in the areas of Quality of Experience (QoE) modeling and Adaptive Bitrate (ABR) algorithms to meet this demand. While learningbased approaches have demonstrated substantial progress using large-scale datasets, existing QoE models often focus on systemlevel metrics such as bitrate and resolution within the playback buffer, neglecting the quality as perceived by the human eye. Simultaneously, many learning-based ABR algorithms exhibit limited robustness in dynamic environments due to their reliance on a one-size-fits-all strategy, which fails to adapt effectively to complex, real-world conditions. In this paper, we propose an integrated system that addresses these limitations by combining an accurate QoE model with an environment-robust adaptation algorithm to enhance user satisfaction in diverse and dynamic environments. First, we introduce RetQoE, a novel approach that accurately estimates the user’s actual QoE by focusing on the quality of video content as perceived by the viewer, rather than on conventional system metrics. Then, we design PVA, a meta-reinforcement learning-based adaptation that rapidly adjusts its policy to varying environments. We systematically integrate RetQoE and PVA, enabling PVA to update its policy with feedback from RetQoE in just a few steps online. We demonstrate the effectiveness of RetQoE+PVA through extensive evaluations in diverse environments, outperforming conventional learning-based algorithms across various metrics.

Deploying high-spec cameras in video systems often falls short of user expectations. Leveraging advancements in deep learning, we propose a mobile, lightweight, real-time video enhancement system. Our approach adopts cutting-edge models and introduces novel optimization techniques for real-time streaming, improving low-resolution, grayscale, and low frame-rate videos. Preliminary evaluations show significant improvements in PSNR and SSIM, while visual assessments confirm substantial quality enhancements while maintaining real-time processing requirements.

The shift from traditional PCs and TVs to mobile devices such as smartphones and tablets has significantly transformed the video streaming landscape. Traditional Quality of Experience (QoE) models, predominantly designed for larger screens, fall short in addressing the nuances of mobile consumption, often misguiding bandwidth usage and quality delivery. This paper introduces a novel user-oriented QoE model tailored for the mobile environment, which accounts for heterogeneous viewing environments. Unlike conventional models that estimate QoE based solely on bitrate and resolution, our approach encompasses the entire video streaming pipeline, from server transmission to the user's perception. In addition, we design lightweight but effective QoE models for mobile devices. This work bridges the gap between user experience and QoE modeling, offering a path toward more adaptive and efficient video streaming services for the increasingly mobile-centric world.

The prevalence of video streaming applications has led to an escalation in users' demands for high-quality services. Numerous endeavors have been undertaken in the realm of quality-of-experience (QoE) models and adaptive bitrate (ABR) algorithms to fulfill this demand. Nevertheless, the existing QoE models exhibit a significant gap with users' actual experience. ABR algorithms are vulnerable in dynamic network environments. We present an integrated system with an accurate QoE model and an environment-robust adaptation algorithm to ensure high user satisfaction in dynamic network conditions. We define a QoE model that accurately estimates the user's QoE by considering the viewing environment and video content. We then design a meta-reinforcement learning-based adaptation algorithm that adapts to dynamic network conditions. We systematically integrate them, allowing it to update its policy with QoE feedback within a few shots.