Video Quality Assessment System at Youku

Jing Li (Alibaba Group, China)

With the fast proliferation of online video sites and social media platforms, user, professionally and occupationally generated content (UGC, PGC, OGC) videos are streamed and explosively shared over the Internet. Consequently, it is urgent to monitor the content quality of these Internet videos to guarantee the user experience.  Youku, a leading Chinese video hosting service platform, has to establish a video quality assessment system to evaluate the whole life cycle of a video from capturing, production, post-processing, to encoding, transmission and delivering to users. Video formats at Youku include UltraHD, HDR, AR, Free View-point video, VR, etc.  In this talk, the general subjective quality assessment system will be introduced for different application purposes. Then, the corresponding objective quality metrics and the guarantees of the users' experience will be illustrated for better understanding.


Jing Li received her BSc degree from Xi’an University of Technology, China in 2006, M.S. degree in Pattern Recognition and Intelligence Systems from Xidian University, China in 2010, and Ph.D. degree in computer science from the Image and Video Communications (IVC/IRCCyN) Group in Polytech Nantes, University of Nantes, France in 2013. She is now with Alibaba Group as a senior staff algorithm engineer and Head of Moku laboratory since 2019, leading teams working on computer vision, video enhancement, video quality assessment, and VR/AR/FVV interactive technology in immersive multimedia. From 2013 to 2019, she has been working in IPI/LS2N lab as a researcher. From 2014 to 2016, she was also an assistant professor at the University of Nantes, France. Currently she is a member of Video Quality Experts Group (VQEG) and IEEE Standard working group P3333.1. She is a contributor of International Telecommunication Union (ITU) Standards ITU-T Rec. P.914, P.915 and P.916, and IEEE Standard P3333.1.1. Her research interests include image quality assessment, QoE of immersive multimedia including both psychophysical study and objective modeling, deep learning, active learning and data mining.