Research Fellow at Media Lab
Dr. Meng-Ping Kao received his Ph.D. from the Electrical and Computer Engineering Dept., University of California, San Diego in 2008. His research interests include video processing and compression, in particular Scalable Video Coding (SVC). He published 10+ IEEE journal and conference paper in the fields of scalable motion vector, residuals, and bitstream selector.Dr. Kao joined Qualcomm Inc., San Diego, US in 2008. He was responsible for designing the H.264 encoder IP inside the Qualcomm mobile SoC chipset. Dr. Kao joined Apple Inc., Cupertino, US in 2012. He was responsible for FaceTime real-time video codec and iTunes high-quality video service system. The iTunes distributed computing architecture and framework was successfully brought to market under his leadership. In 2018, Dr. Kao joined Tencent Media Lab as a Research Fellow. He built and led the artificial intelligence, video processing, compression, quality evaluation, and cloud service teams. In Dec. 2018, Tencent Liying, a brand new high-quality human perceptual video service platform, was debuted under his leadership. His vision of the next generation of multimedia revolution is comprised of innovations that are made possible by artificial intelligence, big data, and cloud service.
Tracks he partipates
Multimedia QoE is very rich, the system of live and VOD is different, the system of long video and short video is different, and the system of mobile, PC and television are also different. The multimedia product will bring different UE from different periods including produce, transcode and delivery. The infinite topic is how to measure, track and change. The topic will share and understand various of UE according to different angles.
Non-reference (NR) quality assessment plays an important role in those practical applications where golden references are not available. In this talk, the actual research and development experience of Tencent Liying is used as an example to demonstrate the benefits of a NR quality metric. When used properly, a good NR quality metric can guide the strength of a video enhancement algorithm, to effectively achieve better human perception.For those algorithm developers who work on image or video enhancement based on human perception, this talk shall serve as a catalyst for further innovative brain storming. We hope to see more and more NR quality assessment applications in the field of human perceptual multimedia industry.