| Keynote Speaker Ⅰ |
| |
 |
| Xin Yao |
| Lingnan University, China |
| IEEE Fellow |
| |
| Brief Introduction: Xin Yao is the Vice President (Research and Innovation) and Tong Tin Sun Chair Professor of Machine Learning at Lingnan University, Hong Kong, China. He is a Fellow of IEEE and Hong Kong Academy of Engineering. He served as the President (2014-15) of IEEE Computational Intelligence Society (CIS) and the Editor-in-Chief (2003-08) of IEEE Transactions on Evolutionary Computation. His major research interests include evolutionary computation, machine learning and AI ethics. His work won the 2001 IEEE Donald G. Fink Prize Paper Award; 2010, 2016 and 2017 IEEE Transactions on Evolutionary Computation Outstanding Paper Awards; 2011 IEEE Transactions on Neural Networks Outstanding Paper Award; 2010 BT Gordon Radley Award for Best Author of Innovation (Finalist); and other best paper awards at conferences. He received the 2012 Royal Society Wolfson Research Merit Award, 2013 IEEE CIS Evolutionary Computation Pioneer Award, and the 2020 IEEE Frank Rosenblatt Award. |
| |
| Speech Title: From Distributive Fairness to Procedural Fairness in Machine Learning |
| |
| Abstract: Distributive fairness in machine learning focuses on fairness of machine learning model outcomes. There have been many metrics proposed in the literature in measuring different aspects of distributive fairness, e.g., statistical parity, equal opportunity, predictive rate parity, etc. However, there are inherent conflicts among some of these metrics. There are also inherent conflicts between fairness and model accuracy. This talk will first introduce a multi-objective learning approach to training fairer machine learning models, where the set of learning objectives may change adaptively during training. Then the talk will present our latest work on procedural fairness, where we study how the decision process of a machine learning model should be fair and unbiased against sensitive groups. In our work, we use explainable artificial intelligence (XAI) technique, namely feature attribution explanation (FAE), to capture the decision process of ML models. We proposed a novel metric to evaluate the group procedural fairness of ML models, called GPFFAE. We validate the effectiveness of GPFFAE on a synthetic dataset and eight real-world datasets. Our experimental studies have revealed the relationship between procedural and distributive fairness of ML models. After validating the proposed metric for assessing the procedural fairness of ML models, we then propose a method for identifying the features that lead to the procedural unfairness of the model and propose two methods to improve procedural fairness based on the identified unfair features. Our experimental results demonstrate that we can accurately identify the features that lead to procedural unfairness in an ML model, and both of our proposed methods can significantly improve procedural fairness while also improving distributive fairness, with a slight sacrifice on the model performance. |
| |
| Keynote Speaker Ⅱ |
| |
 |
| Hesheng Wang |
| Shanghai Jiao Tong University, China |
| |
| Brief Introduction: Dr. Hesheng Wang, Tang Junyuan Distinguished Professor at Shanghai Jiao Tong University and Dean of Pujiang International College, serves as Vice Chairman of the Mixed Intelligence Technical Committee of the Chinese Association of Automation and Vice Chairman of the Intelligent Vehicle and Robotics Branch of the China Instrument and Control Society. He currently or previously held editorial board positions for international journals including TRO, TMECH, TASE, RAL, IJHR and RIA, and has worked as Senior Editor of IEEE/ASME Transactions on Mechatronics, Advisory Editorial Board Member of Advanced Intelligent Systems, and Editor-in-Chief of Robot Learning. As the principal investigator, he has led numerous research projects such as the National Key R&D Program of China, the National Science Fund for Distinguished Young Scholars, the National Science Fund for Excellent Young Scholars, and Key Projects of the NSFC Joint Fund. His academic honors include the National Baosteel Outstanding Teacher Award, Shanghai Rising-Star Program for Young Scientists, and Shanghai Dawn Program. He acted as General Chair of the top robotics conference IROS 2025, and also served as General Chair for RCAR 2016 and ROBIO 2022. |
| |
| Speech Title: Robotic Visual Navigation and Manipulation |
| |
| Abstract: This report focuses on the two core functions of service robots, namely mobility and manipulation. It begins by outlining the current status of the industry and technological development, as well as the challenges encountered. Subsequently, it presents the key achievements made by the team through long-term research into the core technical issues in smobility and manipulation: To address navigation in dynamic environments, a series of localization and navigation methods have been proposed, enabling robust perception and localization via visual fusion for mobile robots in large and complex scenarios. For robotic manipulation in complex environments, an adaptive visual servoing framework and learning-based mobile manipulation methods have been developed. Furthermore, a versatile vision-based methodology has been established for mobile manipulation, advancing the key generic technologies of service robots. |
| |
| Keynote Speaker Ⅲ |
| |
 |
| Chao Shen |
| Xi'an Jiaotong University, China |
| IEEE Fellow |
| |
| Keynote Speaker Ⅳ |
| |
 |
| Zhongkui Li |
| Peking University, China |
| |
| Keynote Speaker Ⅴ |
| |
 |
| Tao Yang |
| Northeastern University, China |
| |
Brief Introduction: Tao Yang is a Professor and Ph.D. Supervisor at Northeastern University, China. He is a National High-Level Young Talent, a Distinguished Professor under the Changjiang Scholars Program, and an IET Fellow. His research primarily focuses on distributed cooperative control and optimization, industrial artificial intelligence, and the integration of intelligent optimization and control. Prof. Yang has led numerous Key Projects and Major Program subprojects funded by the National Natural Science Foundation of China, as well as projects under the National Key Research and Development Program of China. He has published more than 100 journal papers, including over 50 papers in IEEE Transactions and IFAC journals. He received the Second Prize of the National Teaching Achievement Award in Higher Education (Graduate Education) in 2022 (ranked 4th among the 5 contributors) and the Second Prize of the Natural Science Award from the Chinese Association of Automation in 2023 (ranked 1st among the 5 contributors). Prof. Yang currently serves as an Associate Editor of Acta Automatica Sinica and an Editorial Board Member of IEEE TCST, IEEE TCNS, and IEEE TNNLS. He is also the Chair of the Technical Committee on Big Data of the Chinese Association of Automation and a member of multiple technical committees of IEEE CSS, IEEE IES, and IFAC.
|
| |
| Speech Title: Recent Advances in Distributed Nonconvex Optimization |
| |
| Abstract: Distributed optimization algorithms solve large-scale optimization problems through cooperation and coordination among multiple agents. Compared with traditional centralized optimization approaches, distributed optimization algorithms provide greater flexibility, scalability, and computational efficiency. However, most existing distributed optimization algorithms do not account for the nonconvexity of agents' objective functions. Moreover, they often require continuous communication among agents and do not consider the practical limitation of limited communication channel capacity. To address the nonconvexity of objective functions, a distributed first-order primal-dual algorithm for nonconvex optimization is proposed. To eliminate the need for continuous communication among agents, an event-triggered distributed nonconvex optimization algorithm is developed. To cope with the limitation of finite communication channel capacity, a quantized communication-based distributed nonconvex optimization algorithm is further proposed. The convergence of all the proposed algorithms is rigorously established through theoretical analysis, and their effectiveness is demonstrated through simulation results. |
