Admin Login
Username:

Password:

Introduction

I am an academic staff member of Machine Learning and Optimization Research Group in School of Computer Science at The University of Manchester. I do research and teaching in computer science and computational cognitive systems. My Research Interests mainly lie in machine learning, pattern recognition, machine perception, computational cognitive modelling and their applications in intelligent system development. I have been in the current post since 2003. I was with Peking University, The University of Birmingham, The Ohio State University, Kyushu Institute of Technology and Tsinghua University. I was a visiting senior researcher at Microsoft Research Asia and a visiting professor at The Hong Kong Polytechnic University. At present, I am a senior member of IEEE, a member of IEEE Computational Intelligence Society, a member of International Neural Network Society and a member of British Computer Socity. I have served as an associate editor or a member of editorial board for a number of academic journals including Neural Networks: An Official Journal of JNNS, ENNS and JNNS, and IEEE Transactions on Neural Networks. For all my academic activities, see my Academic Activity page for details.

I am keen having self-motivated PhD students join my Machine Learning and Perception (MLP) Lab.  For prospective applicants who have research interests similar to mine (for details, see my Research Interest and Publication pages) , they are strongly suggested visiting my Research Student page first prior to contacting me, where you should be able to find a proper academic inquiry procedure and important information including my PhD Research Project Suggestions, our PhD application procedure and the funding information for EU and non-EU applicants. For a junior researcher who recently received a PhD in a discipline relating to my research interests, they can directly contact me for the further information if they want to visit my MLP Lab.


Spotlighted Researches in MLP Lab (2015-Present)

Our work on Feature Distribution Learning has yielded two publications. The work on Theoretical Aspects of Sparse Filtering has been published in Neural Networks and its pre-print is availalbe from HERE. The Covariant Adaptation via Sparse Filtering paper is available from HERE.

Our project on Zero-shot Visual Recognition has yielded a number of publications. The work on Bidirectional Latent Embedding Learning has been published in International Journal of Computer Vision (the pre-print is availalbe from HERE and the source code in Matlab is available on our Project Website). The paper on Multi-label Zero-shot Human Action Recognition is availalbe from HERE  (the source code in Matlab is available on our Project Website). The work on Alternative Semantic Representations for Zero-shot Human Action Recognition was presented in ECML'17, and the paper and source code are available on our Project Website.

We recently proposed a novel procedural content generation framework based on machine learning, Learning-Based Procedural Content Generation (LBPCG), to generate quality game content and the personalised video games. The work has been publisehd in IEEE Transactions on Computational Intelligent and AI Games and the preprint is available from HERE. If you are interested in playing a personalised first-person shooter game generated by our prototype, please visit the LBPCG-Quake page. Under the LBPCG framework, we have further developed Constructive Primitives via Learning for Super Mario Bros and their applications to ON-LINE and ADAPTIVE (PERSONALISED) Super Mario level generation. To play on-line/personalised Super Mario games and see the technical details including source code and visualisation tools, go to our Learning Constructive Primitives of Super Mario Games Website.

Our work in Learning Contextualized Semantics from Co-occuring Descriptive Terms has been published in Neural Networks. The pre-print is availalbe from HERE  and the source code in Matlab is available upon request and agreement. This work has led to non-trivial applications, e.g., Multi-label Zero-shot Learning (see our Technical Report for details) and Music Information Retrieval (published in ACM Transactions on Intelligent Systems and Technology and the pre-print is availalbe from HERE).

Our work in Spatial Cognition Modelling has made substatial progresses, which leads to a number of papers published in PLoS One, Neural Networks, Cognitive System Research and Biologically Inspired Cognitive Architectures etc. Those pre-prints are available on Spatial Cognition Modelling Publications.



This page was last updated at 04:44pm, November 17th, 2017.
Since 1st August 2014, 431369 visitors have visited this website.