About Me 

I am an Assistant Professor in the School of Science and Engineering at The Chinese University of Hong Kong, Shenzhen.

I received my bachelor’s degree in pure and applied mathematics from the Department of Mathematics at Sun Yat-sen University in Guangzhou, China. I subsequently pursued systematic training in mathematics in the United States and Canada, earning three master’s degrees spanning pure, applied, and computational mathematics. My studies placed particular emphasis on analysis, differential geometry, and mathematical optimization. My mathematical training was strongly shaped by mentors working in the traditions of geometric analysis and geometric control theory, whose research tastes continue to influence how I formulate research questions and develop methods in artificial intelligence.

Following my mathematical training, I chose to return to China and spent several years dedicating my career at the forefront of tech industrial application. At Baidu (Beijing Headquarters), I served as an algorithm engineer on the L3 autonomous-driving team within the Intelligent Driving Group, where I worked alongside Dr. Zeng Wang and Dr. Liyao Tang on trajectory prediction and spatiotemporal traffic modeling for the Baidu Apollo platform. I later joined WeBank’s AI Department in Shenzhen, where I worked at the intersection of industrial R&D and academic research. Together with Dr. Lixin Fan, Professor Qiang Yang, Dr. Dashan Gao, Dr. Kam Woh Ng, Dr. Ben Tan, and Professor Yang Liu, I conducted research on federated learning, federated transfer learning, and privacy-preserving machine learning for healthcare data. During this period, in my role at WeBank’s AI Department, I also collaborated with Ruihui Zhao and Professor Yefeng Zheng at Tencent Healthcare’s Jarvis Research Center as part of an institutional collaboration between WeBank and Tencent on federated learning for healthcare.

I subsequently left industry to pursue a Ph.D. in Computer Science at the College of Computing and Data Science at Nanyang Technological University in Singapore, under the supervision of Professor Cuntai Guan. My doctoral research focused on geometric deep learning methods for brain–computer interfaces, particularly covariance-based EEG decoding. During my doctoral studies, I was invited to undertake a research visit to the Advanced Telecommunications Research Institute International (ATR) and RIKEN in Japan, where I worked with Dr. Reinmar J. Kobler and Dr. Motoaki Kawanabe on a geometric deep learning approach to integrating simultaneous EEG–fMRI data. My doctoral dissertation, Geometric Methods for Covariance-Based Neural Decoding, received an Honorable Mention for the College’s Outstanding Ph.D. Thesis Award.

After completing my Ph.D., I conducted postdoctoral research with the MIND team, jointly supported by Inria Saclay–Île-de-France, CEA, and Université Paris-Saclay, under the supervision of Dr. Bertrand Thirion. During this period, I broadened my research on geometric methods beyond neural decoding to neuroimaging, including symmetric positive-definite matrix learning, generative modeling of functional brain connectivity, and cross-dataset benchmarking. I also collaborated with Dr. Bruno Aristimunha, Dr. Reinmar J. Kobler, Dr. Antoine Collas, Professor Florent Bouchard, and Professor Sylvain Chevallier to develop open-source software for geometric deep learning and reproducible research tools for neural decoding and neuroimaging analysis.

My work has appeared in journals and conferences including IEEE TPAMI, IEEE TNNLS, Artificial Intelligence (AIJ), ICLR, NeurIPS, and ICML, with one paper selected for an ICLR Spotlight presentation. I currently serve as an Associate Editor of IEEE Transactions on Medical Robotics and Bionics and as a Recommender for the PCI StatML Community. I also regularly review submissions for more than ten international journals and major conferences.

我现任香港中文大学（深圳）理工学院助理教授。

我于广州中山大学数学系获得纯数学与应用数学学士学位。此后，我先后在美国和加拿大接受系统的数学训练，并获得三个硕士学位，专业方向涵盖纯数学、应用数学和计算数学。我的学习重点包括分析学、微分几何与数学最优化。我的数学训练深受几何分析与几何控制论学术传统中多位导师的影响；他们的研究品味至今仍在影响着我在人工智能研究中提出问题和构建方法的方式。

完成数学训练后，我选择回国，并在科技产业应用的最前沿全职工作数年。在百度（北京总部）期间，我担任智能驾驶事业群 L3 自动驾驶团队的算法工程师，主要从事百度 Apollo 平台的轨迹预测与时空交通建模, 并与王政博士和汤力遥博士等共事。之后，我加入位于深圳的微众银行人工智能部门，与首席科学家范力欣博士、加拿大两院院士杨强教授、高大山博士、吴锦和博士、谭奔博士和刘洋教授等共事，开展面向医疗数据的联邦学习、联邦迁移学习和隐私保护机器学习研究，并参与医疗人工智能研究与产学研合作。在此期间，我还以微众银行人工智能部门研究人员的身份，参与微众银行与腾讯（深圳总部）之间的机构合作，与腾讯医疗天衍实验室的赵瑞辉研究员和郑冶枫教授共同开展面向医疗健康的联邦学习研究。

此后，我离开产业界，赴新加坡南洋理工大学计算与数据科学学院攻读计算机科学博士学位，由新加坡工程院院士关存太教授指导。博士期间，我主要研究面向脑机接口的几何深度学习方法，尤其关注基于协方差表示的脑电解码。在攻读博士学位期间，我受邀赴日本国际电气通信基础技术研究所（ATR）和理化学研究所（RIKEN）开展访问研究，与 Reinmar J. Kobler 博士和 Motoaki Kawanabe 博士合作，探索基于几何深度学习的同步脑电—功能磁共振成像数据融合方法。我的博士论文题为 Geometric Methods for Covariance-Based Neural Decoding，获南洋理工大学计算与数据科学学院优秀博士论文奖荣誉提名奖。

博士毕业后，我在法国国家信息与自动化研究所（萨克雷—法兰西岛研究中心）、法国原子能和替代能源委员会及巴黎萨克雷大学共同支持的 MIND 团队从事博士后研究，由法国科学院院士 Bertrand Thirion 博士指导。在此期间，我将几何方法的研究从神经解码进一步拓展至神经影像，相关工作涵盖对称正定矩阵学习、功能性脑连接的生成建模以及跨数据集基准评测。我还与 Bruno Aristimunha 博士、Reinmar J. Kobler 博士、Antoine Collas 博士、Florent Bouchard 教授和 Sylvain Chevallier 教授等合作，共同开发面向神经解码与神经影像分析的几何深度学习开源软件及可复现研究工具。

我的研究成果发表于IEEE TPAMI、IEEE TNNLS、Artificial Intelligence（AIJ）以及 ICLR、NeurIPS 和 ICML 等国际期刊与会议，其中一篇论文入选 ICLR Spotlight。目前，我担任 IEEE Transactions on Medical Robotics and Bionics 副编辑及 PCI StatML Community 推荐编辑，并长期受邀为十余种国际期刊和重要学术会议提供同行评审。