Ma, Siwei is a professor in the School of Computer Science. He obtained his B.Sc. from Shandong Normal University in 1999, and Ph.D. from Institute of Computing Technology, Chinese Academy of Sciences in 2005 respectively. His research interests include video coding, processing, transmission, and quality assessment.
Dr. Ma has published more than 100 research papers, and most of them are published in top-tier conferences and journals, such as IEEE TIP, IEEE TCSVT, IEEE TMM, JSTSP, and DCC. He has served in the Technical Program Committee of various international conferences including ISCAS, ICME, VCIP, PCM, and as co-chair of TPC VCIP2017, Area Chair of ISM2015. He is serving as an associate editor of IEEE TCSVT (2016~), and JVCIR (2014~).
Dr. Ma has more than ten research projects including NSFC, 863 project, etc. His research achievements are summarized as follows:
1) Transform/prediction based high efficiency video coding techniques: He proposed series of high efficiency video coding tools, including rate-distortion optimization transform, adaptive motion vector resolution, rate distortion optimization mode decision, and non-local structure based loop filtering etc. Many of these works have been published on top journals, e.g. IEEE TCSVT and IEEE TIP, and serval techniques have been accepted by HEVC/H.265 and AVS video coding standards.
2) Perceptual quality metric and perceptual distortion optimization coding: Perceptual coding is very promising for future video applications. But the quality metric is an important issue in perceptual coding, and it would affect the design of perceptual coding methods significantly. He has proposed a spatial-temporal structural information based video quality metric, which has lower computational complexity while approximating the human visual testing results accurately. Moreover, based on the proposed perceptual quality metrics, optimized perceptual coding was researched and the proposed perceptual coding method can achieve more 20% bits saving compared with the traditional coding methods.
3) 3D video coding and VR video processing: 3D video and VR applications becomes more and popular. His research interests focus on high efficiency 3D video coding and processing, including multi-view prediction coding, texture plus depth joint optimization coding, and panorama video stitching and coding. He proposed a low complexity synthesis view distortion estimation model, which has been adopted by MPEG 3DV video coding standards. His team has developed an AVS2 based multiview broadcasting system, which can support 8 HD video streams coding simultaneously, and a 4K VR system with real time panorama video stitching and streaming.
