I currently serve as an Associate Professor at the School of Mathematics and Statistics, Central South University (CSU) in Changsha, where I am privileged to explore the fascinating intersections of mathematics, statistics, and machine learning, with a consistent focus on physical artificial intelligence (Physical AI) that engages directly with real world data, systems, and constraints. This Physical AI orientation spans two interconnected fronts: one concerning brain and embodied intelligence, rooted in neural dynamics, and biomechanical signals; the other concerning the global communication environment, rooted in electromagnetic propagation, spatial signal distributions, and large scale wireless networks. Before joining CSU, I was a Postdoctoral Researcher at The Chinese University of Hong Kong, Shenzhen. My academic foundation was built through dual PhD programs at the Chinese Academy of Sciences and Radboud University in the Netherlands, where I cultivated a deep appreciation for interdisciplinary research and collaboration.

本人拥有丰富的企业工作经历，企业经验覆盖智能化技术的研发、产品交付与工程管理，涉及信息技术（含软硬件）、智能装备、互联网及机器人等行业。同时，现任中南大学本科生与研究生数学建模教练组指导老师，及中南大学卓越工程师学院升华科创班核心导师，具备较深厚的产学研合作经验，主持与领先新能源材料企业及智能网联汽车企业的产学研合作项目，推动前沿技术在产业场景中的落地与转化，对前沿技术落地和复杂工程现实具有独到见解。真正有效的智能方法必须植根于物理世界的环境、信号、结构和动态演化过程，而非仅依赖理想化的数据假设和数学推导。本课题组的物理人工智能研究主要关注两个层面的真实挑战：一是人脑与具身系统的生物物理约束，二是低空环境的电磁物理规律与空间异质性。

Research Focus

My research centers on addressing fundamental challenges in Bayesian deep learning and complex multimodal signal processing, with a particular emphasis on developing innovative methodologies that are grounded in these two physical AI fronts and deployable under practical engineering constraints. Key areas of interest include:

• High-Dimensional Statistics & Manifold Representation Learning: Developing statistical methods, tools, softwares, and agents for online statistical analysis, manifold structure discovery, and automated inference on ultra-high-dimensional data, with applications in revealing complex patterns in domains such as materials science and biomedical data. By explicitly modeling the underlying geometric and physical regularities of these data, our approaches bridge high dimensional theory with tangible material behaviors and biological signals. Related works have been published in Annual Meeting of the Association for Computational Linguistics (ACL), Neural Computing and Applications, and Computational Materials Science. 本团队自主开发了具有自主知识产权的深度统计计算软件及在线引擎。

• Multimodal & Scalable Gaussian Processes: Developing advanced probabilistic models through deep architectures and kernel methods to achieve robust, interpretable, and flexible learning, with particular applications in decoding complex brain signals and modeling underlying brain-computer systems (脑机接口). These models are designed to capture the nonstationary, nonlinear dynamics of neural signals in real time, enabling reliable inference in physical human machine closed loops. Related works have been published in IEEE Transactions on Neural Networks and Learning Systems (TNNLS), Machine Learning (ML), Pattern Recognition (PR), Signal Processing, the European Conference on Machine Learning (ECML), The Genetic and Evolutionary Computation Conference (GECCO), and IEEE Transactions on Biomedical Engineering (TBME, primarily contributed by graduate students). 本团队拥有可穿戴式脑电头环二次开发实验系统、脑控机器人系统及底层信号处理模块开发组件。

• Bayesian Multitask Deep Learning & VLMM: Designing principled frameworks that enable efficient knowledge sharing and rapid generalization across diverse tasks, while explicitly accounting for the underlying physical propagation laws and environmental heterogeneity. A key application is in 6G mobile communications for radio map generation (无线通信) and spatial intelligence (空间智能), where these techniques allow models to efficiently construct and predict large-scale spatial signal distributions by leveraging data across multiple scenarios, frequencies, and deployment environments, directly informed by electromagnetic physics and site specific geometries. Related works have been published in IEEE Transactions on Neural Networks and Learning Systems (TNNLS), the International Conference on Acoustics, Speech, and Signal Processing (ICASSP), the IEEE Global Communications Conference (GLOBECOM, primarily contributed by graduate students), and IEEE International Conference on Computer Communications (INFOCOM, primarily contributed by graduate students). 本团队拥有无线电信号感知（5G、6G）采集与生成系统。

Research Projects

I have had the privilege of leading and contributing to several impactful research initiatives, including:

• National Natural Science Foundation of China (Youth Program): Research on Coupled Gaussian Process Regression Networks for Multitask Learning (2021–2022, Principal Investigator).

• Hunan Provincial Natural Science Foundation (Youth Program): Data-Driven Non-Stationary Gaussian Process Learning Models and Algorithms (2023–2026, Principal Investigator).

• China Postdoctoral Science Foundation (General Program):Multi-Output Gaussian Processes for Spatiotemporal Collaboration in IoT Multi-Sensor Systems (2020–2022, Principal Investigator).

• National Key R&D Program of China:Data-Driven and AI-Based Evolution of Future Networks (2019–2023, Team Member).