Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Learn about MARC, a motion-aware rate control framework designed to optimize mobile e-commerce cloud rendering experiences in this 15-minute conference presentation from USENIX ATC '25. Discover how researchers from Alibaba Group and University of Chinese Academy of Sciences addressed the challenge of latency spikes during peak user engagement periods that lead to early session abandonment in 3D shopping environments. Explore the framework's innovative approach to aligning bitrate decisions with real-time user Quality of Experience (QoE) preferences by setting dynamic QoE objectives based on actual user engagement behavior. Understand how MARC differentiates between motion and non-motion frames to capture varying latency and quality requirements, while employing stochastic optimization to maximize overall user experience. Examine the impressive results from extensive deployment across over 1 million user sessions, which demonstrated a 71% reduction in session freeze rates and a 20% increase in user interaction time, significantly enhancing user engagement for e-commerce cloud rendering applications.