Prof. Ce ZHU, University of Electronic Science & Technology of China
Prof. Ce ZHU, University of Electronic Science & Technology of China Ce Zhu is currently a Professor with the School of Information and Communication Engineering, University of Electronic Science and Technology of China, Chengdu, China. His research interests include image/video coding and communications, video analysis and processing, 3D video, visual perception and applications. He has served on the editorial boards of a few journals, including as an Associate Editor of IEEE Transactions on Image Processing, IEEE Transactions on Circuits and Systems for Video Technology, IEEE Transactions on Broadcasting, and IEEE Signal Processing Letters. He has served on technical committees, organizing committees and as track/area/session chairs for over 60 international conferences, including serving as a Technical Program Co-Chair of IEEE ICME 2017. He is a Fellow of the IEEE and a Fellow of the IET. For more information, please visit his homepage at http://www.avc2-lab.net/~eczhu
Speech Title: Deep Learning for In-Loop Filter Design in HEVC
In-loop filters have been employed to reduce coding artifacts in the recent video coding standards including the latest High Efficiency Video Coding (HEVC) standard. However, in the existing approaches, an in-loop filter is always applied to each single frame, without exploiting the content correlation among multiple frames. In the talk, we introduce a multiframe in-loop filter (MIF) for HEVC, which enhances the visual quality of each encoded frame by leveraging its adjacent frames. Extensive experiments show that the MIF approach achieves 11.621% saving of the Bjøntegaard delta bit-rate (BD-BR) on the test set, significantly outperforming the standard in-loop filter in HEVC and other state-of-the-art approaches.
The talk is based on our recent work entitled “A Deep Learning Approach for Multi-Frame In-Loop Filter of HEVC” in IEEE Transactions on Image Processing (in press, https://github.com/tianyili2017/HIF-Database).
Prof. Jianru Xue, Xi'an Jiaotong University, China
Jianru Xue, Phd, Professor, Changjiang Scholar. He got his MS and PhD degrees from Xi’an Jiaotong University in 1999 and 2003, respectively. Since 1999, he joined the Institute of Artificial Intelligence and Robotics at Xi’an Jiaotong University, Xi’an, China, where he currently is a full professor. He had worked in FujiXerox, Tokyo, Japan, from 2002 to 2003, and visited University of California, Los Angeles, from 2008 to 2009. His research interests include computer vision and pattern recognition, machine learning, and autonomous driving. He and his team won the IEEE ITSS Institute Lead Award in 2014. He and his students won the best application paper award in Asian Conference on Computer Vision 2012. He is co-author of the book Statistical Learning and Pattern Analysis approaches to Image and Video Processing，published by Springer-verlag in 2009. He has published 100+ papers in top cited journals and conferences including IEEE TPAMI, IEEE TIP, IEEE TSMCB, ICCV, ECCV, ACM MM, ICPR, etc. He had severed as organization chair or co-chair of several international conferences including VALSE2012, VLPR2011, VLPR2010, ACCV2010, VSMM2006, and so on. He also served on the technical program committee of peer-reviewed conferences ICME, ACCV, ICPR, IVS, etc. More information about Jianru Xue can be found in his personal web pages: http://gr.xjtu.edu.cn/web/jrxue
Speech Title: Hybrid Augmented Intelligent Driving Vehicles
Autonomous driving is an undoubtedly disruptive technology in the field of artificial intelligence, which will affect almost all aspects of our communities. However, making a self-drivng car capable of autonomous intelligence in real traffics still faces many open and challenging problems. Thus, we need to adopt the role of human driver or learning from skilled human drivers. This line of thoughts leads to an emerging research topic, hybrid augmented intelligent. In this report, we discuss the hybrid augmented intelligent driving systems which can integrates human intelligence and machine intelligence, and its core ideas are inspired by both brain and neuro science. We also report our recent research results on robot learning methods including Learning by demonstration, imitation learning, and interactive learning.
Prof. Weiwei Xu,
State Key Lab of CAD&CG, Zhejiang University, China
Weiwei Xu is now a research professor at State Key Lab of CAD&CG, Zhejiang university. He was a Post-Doc in Ritsumeikan university in Japan, Researcher in Microsoft research asia, Qianjian distinguished professor in Hangzhou normal university. His research direction includes 3D sensing, deep learning, physics-based simulation and 3D printing. He has published more than 80 papers on prestigious conferences and journals, such as ACM Trans. On Graphics, CVPR, IEEE TVCG and Computer Graphics Forum. He won the outstanding your researcher award from NSFC in 2014, and now is the PI of NSFC Key Project on data-driven 3D modeling.
Speech Title: Deep Feature learning for Computer Graphics
AI-drive graphics is a multi-disciplinary research
direction. It focuses on the problem of learning features and patterns
from the large amount of graphics and image data, and representing the
semantic constraints in the feature space to simplify the algorithm,
which can significantly improve the algorithm speed and the quality of
results. Since the deep learning technology provides us the powerful
learning framework, it becomes popular in computer graphics nowadays.
This talk is on how to apply deep learning algorithms to image
understanding and 3D shape editing. We will report how to design
network structure and control the gradient flow in the network to
achieve high-quality results.