Invited Speakers
Invited Speeches

Prof. Le Wu

Hefei University of Technology, China

Dr. Le Wu is a professor at Hefei University of Technology, China. She received the Ph.D. degree from the University of Science and Technology of China. Her current research interests focus on user-centric intelligent systems, such as recommendation, user modeling, and fairness and explainability issues for user-centric applications. She has published 50+ academic papers in leading journals (e.g., TKDE, TOIS, TNNLS, TMSC, TIST, TMM) and conferences (e.g., SIGIR, WWW, AAAI, IJCAI, KDD, MM, SDM, CIKM). She was honored to receive the Distinguished Dissertation Award from China Association of Artificial Intelligence (CAAI) in 2017, and the paper awards from SIAM International Conference on Data Mining (“Best of SDM 2015”). She regularly serves as a reviewer or (senior) PC member for top journals and conferences.

Speech Title: "Fairness in Recommendation: Challenges, Solutions, and Future Directions"

Abstract: Personalized recommendation is a human-centric AI application that plays a pivotal role in influencing user decisions. However, recent research has highlighted the potential for AI systems to introduce discrimination and unfairness. In this talk, we first give an overview of the background of fairness research in AI and recommendation. We then briefly discuss the progress and challenges in fairness research, followed by our recent studies of fairness modeling for recommendation. Finally, we conclude this talk by highlighting potential future directions for research in this area.

Assoc. Prof. Wei Liu

Shanghai Jiao Tong University, China

Wei Liu received the Ph.D. degree from the Department of Automation, Shanghai Jiao Tong University. He was a postdoctoral research fellow of The University of Adelaide and The University of Hong Kong. He is currently an associate professor with the Institute of Image Processing and Pattern Recognition, Shanghai Jiao Tong University, China. His major research interests are robust and fast image filtering and the applications in computer vision and graphics. He has published several papers in major journals and conferences including TPAMI/TOG/TIP/TMM/TCSVT and ICCV/AAAI/IJCAI/IROS.

Speech Title: "Fast and Robust Image Filtering: Methods and Applications"

Abstract: In this talk, I will introduce our recent work on fast and robust image filtering. For fast image filtering, I will talk about our iterative least squares (ILS) framework for fast global image smoothing. Our ILS is highly parallel which can be accelerated through either multi-thread computing and GPU acceleration. It is able to process high-resolution images (e.g. 1080p) at a real-time processing rate on a GPU. In addition, our method can also yield high smoothing quality which is comparable to the state-of-the-art approaches but at a much lower computational cost. For robust image filtering, I will present the generalized framework for edge-preserving and structure-preserving image smoothing proposed by us. Our method is able to achieve various smoothing behaviors and the smoothing property that is not able to be achieved by previous approaches. It is thus applicable to many tasks and the stat-of-the-art performance is also achieved.

Assoc. Prof. Qiufeng Wang

Xi’an Jiaotong-Liverpool University, China

Dr. Qiufeng Wang is currently an associate professor and the head of Department of Intelligent Science at School of Advanced Technology in Xi’an Jiaotong-Liverpool University (XJTLU). He received the Ph.D degree in Pattern Recognition and Intelligence Systems from Institute of Automation, Chinese Academy of Sciences (CASIA) in July 2012, and won Presidential Scholarship of Chinese Academy of Sciences. During 2012 to 2013, he worked as an Assistant Professor at the National Laboratory of Pattern Recognition (NLPR) in CASIA. During 2013 to 2017, he worked at Microsoft and joined in XJTLU in Feb. 2017.  His research interests include pattern recognition and machine learning, specially document analysis and recognition. Dr. Wang has published 50+ papers, including IEEE T-PAMI, ICCV, ICML, and published one book about deep learning in Springer. His research has been supported by both government and industry, including NSFC young programme, NSFC general programme, and CCF-Tencent.

Speech Title: "Towards Long-Tailed Oracle Character Recognition"

Abstract: Deciphering oracle bone script is of great significance to the study of ancient Chinese culture as well as archaeology. Although recent studies on oracle character recognition have made substantial progress with the development of AI techniques, they still suffer from the long-tailed data distribution that results in a noticeable performance drop on the tail classes. In this talk, we will first give a general introduction about the progress of oracle character recognition, then review the methods for the long-tailed issue in this task, and finally introduce our proposed methods for the long-tailed oracle character recognition based on the data augmentation. In addition, the current recognition performance on the benchmark datasets will be reported and the possible future direction for oracle character recognition will also be discussed.

Assoc. Prof. Weiwei Du

Kyoto Institute of Technology, Japan

WEIWEI DU (Member, IEEE) received the B.S. degree from the Tianjin Institute of Urban Construction, China, in 2002, the master's degree from The University of Aizu, Japan, in 2005, and the Ph.D. degree from Kyushu University, Japan, in 2008. She joined the Kyoto Institute of Technology in 2008, as an Assistant Professor. She visited University of Chicago, USA in 2013 as a visiting scholar. She has been a full associate professor in information and human science with the Kyoto Institute of Technology from 2019. Her research interests are to use visual information to help experts obtain effective information by signal/image processing, pattern recognition, computer vision and deep learning. The following researches are ongoing: (1) Log features extraction for experts in studying wood strength, (2) Image inpainting of long-standing damaged artworks for artists, (3) Detection of lung nodules and classification of their benignancy or malignancy for radiologists, (4) Features extraction from fundus images for ophthalmologists, (5) Construction of a system for effective structure elucidation from botanical ingredient in traditional medicine and (6) Development of plastering robot by motion capture.

Speech Title: "Image Inpainting using Automatic Structure Propagation with Auxiliary Line Construction"

Abstract: Existing image inpainting methods used traditional and deep learning methods to restore a large missing region in the damaged image. This often leads to color discrepancy and blurriness. Pre-processing of prior line detection by user assistance is usually employed to reduce the blurry of center region by segmenting the large region into more minor. However, it operates manually, which is time-consuming. This study introduces a technique to generate two-line types: penetrator and interactor in constructing auxiliary lines as guidance. These lines assist structure propagation established automatically, while the remaining small regions are filled by texture propagation. Experiments on large regular masks demonstrate that this study generates higher-quality results than other methods.

Asst. Prof. Yun Gu

Shanghai Jiao Tong University, China

Yun Gu received the Ph.D. degree from the Department of Biomedical Engineering, Shanghai Jiao Tong University. He is currently an Assistant Professor with the Institute of Medical Robotics and Institute of Image Processing and Pattern Recognition, Shanghai Jiao Tong University, China. He has published in major conferences and journals including MedIA/IEEE TMI/TIP/TNNLS/TBME/JBHI and MICCAI/ICRA/IROS. His major research interests are computer-assisted surgical planning and medical image analysis.

Speech Title: "Visual Learning for Pulmonary Anatomical Parsing and Surgery Planning"

Abstract: Endobronchial intervention is increasingly used as a minimally invasive means for the treatment of pulmonary diseases. This process requires accurate pre-operative diagnosis, planning, and intra-operative guidance for precise treatment. In this talk, we will present our recent works on pulmonary anatomical analysis and surgical navigation driven by clinical-friendly priors.

Assoc. Prof. Congduan Li

Sun Yat-sen University, China

Congduan Li (IEEE Member) received the B.S. degree from the University of Science and Technology Beijing, China, in 2008, the M.S. degree from Northern Arizona University, AZ, USA, in 2011, and the Ph.D. degree from Drexel University, PA, USA, in 2015, respectively, all in Electrical Engineering. From October 2015 to August 2018, he was a Post-Doctoral Research Fellow with the Institute of Network Coding, The Chinese University of Hong Kong and with the Department of Computer Science, City University of Hong Kong. He is currently an Associate Professor with the School of Electronics and Communication Engineering, Sun Yat-sen University, China. His research interests lie in the broad areas related with networks, such as coding, security, wireless, storage, and caching.