BU Research Blog

Thanks to the QR funding from the Faculty of Media and Communication, I recently had the privilege to attend the Computer Graphics International 2017 in Yokohama, Japan. CGI is a prestigious conference in the graphics world for many years and attracts interest from people all over the world. We have one paper presented on the conference, as a full paper, it will be published in the Visual Computer Journal by Springer-Verlag.

We have hosted the CGI2012 at Bournemouth. CGI2017 is held at Keio University, Yokohama chaired by Prof. Issei Fujishiro and Prof. Xiaoyang Mao. The first day (27^th June) events are majorly tutorials and workshops. The shining part is the keynote by Prof. Marina Gavrilova from University of Calgary, discussing how to use machine learning for social data analysis through image and video processing, especially multiple modal biometric data acquisition, feature matching, fusion and recognition. One interesting application is gender prediction using image aesthetics.

The second day’s keynote is “reconstructing reality: from physical world to virtual environments” by Prof. Ming Lin from University of North Carolina. She firstly gave an introduction on the history of VR and AR, then talked about her work on human tissue material property estimation from real surgery videos, virtual dressing, crowd and traffic pattern learning from videos of real world, finally the sound and haptic rendering in VR environment. Daniel Thalmann (chair of the conference) held the panel discussion on the future of machine learning in graphics. There were lots of other paper presentations on the second day covering many topics including image, texture, modelling, rendering, deformation and visualization.

The third day’s keynote is “Studies on Humanlike robots” by Prof. Hiroshi Ishiguro. He showed the latest state-of-the-art robot in Japan, the mimic version of himself. Obviously lots of small human robots have been used in some restaurants to entertain guest during food ordering. My student Tao Jiang presented his paper in the morning session.

I chaired a session on the fourth day on papers working on image & example based modelling. There are also some posters presented during the conference.

This is a very good opportunity to network with top experts in the world. I had good chat with many people such as Prof. Henry Fuchs (Guru in the VR world), Prof. Ming Lin, Prof. Marina Gavrilova, Prof. Kwan-Liu Ma (leading expert in the visualization, chaired the top visualization conferences in the world), Prof. Shimin Hu (Tsinghua University), Prof. Enhu Wu (Chinese Academy of Science), and also some other experts from Taiwan, Singapore, South Korea, USA and China etc. There is also a discussion with a professor from German about collaboration on 3D living environment reconstruction from old photos for dementia patient/old people. This is a very exciting and intensive experience for four full days, sharing different ideas, building links and research collaborations.