This is the web page of the Artificial Intelligence track of the ABCP 2021 Annual Conference, to be held on Friday 2nd and Saturday 3rd July 2021.


Day 1: Friday 2nd July 2021 (Zoom link)

1:00-4:00pm: Keynote speeches, Chair: Professor Yaochu Jin, University of Surrey

  • 1:00-2:00pm:  Creating the theoretical foundation of Artificial Intelligence
    • Professor Bo Zhang, Tsinghua University, China
      show more on/off
      Professor Bo Zhang
      Professor Bo Zhang

      Abstract: so far, there is no accepted theoretical foundation for artificial intelligence (or AI) leading to its slow and tortuous development path. The founders of AI put forward the theory of symbolism, which holds that symbol is the basic unit of cognition, and cognition is the manipulation of symbols, i.e., the first generation of AI. It cannot be a complete theoretical foundation for AI due to the symbol grounding problem. At the same time, there is a competing connectionism camp or sub-symbolism (including deep learning), which holds that sub-symbol is the basic unit of cognition, i.e., the second generation of AI. It cannot be a theoretical foundation for AI as well due to its series of defects.

      In this presentation, we will discuss how to build the theoretical foundation of AI, namely the third generation AI. Its basic idea is to combine the first generation of AI-symbolism with the second generation of AI-connectionism. We expect this combination to lay the foundation for the future AI.

  • 2:00-3:00pm:  Artificial intelligence —— Computing, Algorithm, Interaction
    • Professor Qionghai Dai, Tsinghua University, China
      show more on/off
      Professor Qionghai Dai
      Professor Qionghai Dai

      Qionghai Dai is a full Professor in Tsinghua University, the director of the School of Information Science and Technology, and the director of the Institute of Brain and Cognitive Sciences at Tsinghua University. He is also the chairman of Chinese Association for Artificial intelligence.

      Qionghai’s research centres on the interdisciplinary study of Brain Engineering and the next-generation Artificial Intelligence. He has built up various multi-scale multi-dimensional computational imaging instruments, aiming for the simultaneous multi-scale observation of dynamic structures spanning from organelles, cells, tissue, and organs. By developing advanced imaging techniques for the simultaneous recording of millions of neurons, he tries to understand the structures and mechanisms of entire neural circuits on various tasks at single-cell level in mammalian, which can provide theoretical supports for next-generation neuromorphic computing algorithms (including expression, transform, and rules), as a new pathway from Brain Science to Artificial Intelligence.

      Abstract: In this talk, I will present some of my thoughts on AI. Firstly, AI techniques, especially the deep learning techniques, have made a lot of important progresses in recent years. However, they always require huge computing resources, which gradually becomes a limitation in many applications. Will we have disruptive computing technologies in the future to fully solve this problem? Secondly, the current AI techniques mainly work in specific tasks, which are far from human intelligence. This leads us to think about how to make a breakthrough by the inspirations from brain science and cognitive neuroscience. Thirdly, AI is not an isolated brain. It requires an “AI body” to interact with the environment and humans. Then how to build such an “AI body”? Finally, how do we evaluate AI technologies and what is the nature of intelligence? The earliest Turing test can no longer adapt to the current or the future AI which is highly developed. How to establish a new Turing test is also an interesting question to us.

  • 3:00-4:00pm: Human-artificial intelligence partnerships

4:00-4:10pm: Break

4:10-5:00pm: Short presentations, Chair: Professor Huiru Zheng, Ulster University

  • 4:10-4:20pm: Meeting societal challenges: Big data driven AI-enabled approaches
    • Professor Liangxiu Han, Manchester Metropolitan University
      show more(on/off)

      Professor Liangxiu Han has a PhD in Computer Science from Fudan University, Shanghai, P.R.China (2002). She is currently a Professor of Computer Science at the Department Computing and Mathematics, Manchester Metropolitan University. She is a co-Director of Centre for Advanced Computational Science and Deputy Director of ManMet Crime and WellBeing Big Data Centre. Her research areas mainly lie in the development of novel big data analytics/Machine Learning/AI, and development of novel intelligent architectures that facilitates big data analytics (e.g., parallel and distributed computing, Cloud/Service-oriented computing/data intensive computing) as well as applications in different domains (e.g. Health, Precision Agriculture, Smart Cities, Cyber Security, Energy, etc.) using various large scale datasets such as images, sensor data, network traffic, web/texts and geo-spatial data. As a Principal Investigator (PI) or Co-PI, Prof. Han has been conducting research in relation to big data processing and data mining, cloud computing/parallel and distributed computing (funded by EPSRC, BBSRC, Innovate UK, Horizon 2020, British Council, Royal Society, Industry, Charity, respectively, etc.).

      Prof. Han is a member of EPSRC Peer Review College, an independent expert for Horizon 2020 proposal evaluation/mid-term project review, and British Council Peer Review Panel. She is served as an associate editor/a guest editor for a number of reputable international journals and a chair (or Co-Chair) for organisation of a number of international conferences/workshops in the field. She has been invited to give a number of keynotes and talks on different occasions (including international conferences, national and international institutions/organisations).

      Abstract: By 2025, the total size of digital data generated by social networks, sensors, biomedical imaging and simulation devices, will reach an estimated 163 Zettabytes (e.g. 163 trillion gigabytes) according to IDC report. This type of ‘big data’, together with the advances in information and communication technologies such as data analytics/machine learning/AI, Internet of things (IoT), connected smart objects, wearable technology, ubiquitous computing, is transforming every aspect of modern life and bringing great challenges and spectacular opportunities to fulfil our dream of a sustainable smart society. This talk will focus on new developments and methods on scalable learning from large scale data and present real case studies to demonstrate how we applied big data driven, AI enabled approaches in various application domains such as Health, Food to address society challenges.

  • 4:20-4:30pm: Distributed learning over networks
    • Professor Zhengtao Ding, University of Manchester
      show more(on/off)

      Zhengtao Ding received B.Eng. degree from Tsinghua University, Beijing, China, and M.Sc. degree in systems and control, and the Ph.D. degree in control systems from the University of Manchester Institute of Science and Technology, Manchester, U.K. After working in Singapore for ten years, he joined the University of Manchester in 2003, where he is currently Professor of Control Systems with the Department of Electrical and Electronic Engineering. He is the author of the book: Nonlinear and Adaptive Control Systems (IET, 2013) and has published over 300 research articles. His research interests include nonlinear and adaptive control theory and their applications, more recently network-based control, distributed optimization and distributed learning, with applications to power systems and robotics. Prof. Ding has served as the Subject Chef Editor of Nonlinear Control for Frontiers, and Associate Editor for the IEEE Transactions on Automatic Control, IEEE Control Systems Letters, Transactions of the Institute of Measurement and Control, Control Theory and Technology, Mathematical Problems in Engineering, Unmanned Systems, and the International Journal of Automation and Computing, etc. He is a member of IEEE Technical Committee on Nonlinear Systems and Control, IEEE Technical Committee on Intelligent Control, and IFAC Technical Committee on Adaptive and Learning Systems.

      Abstract: There are many challenges and opportunities, such internet of things, big data, machine learning, smart grid etc in the area of network-connected systems and control applications, in particular, in the areas relating to distributed learning, optimisation, decision making and control. This talk will cover some recent activities in relation to machine learning carried out in the speaker’s group in University of Manchester, in particular, consensus-based distributed machine learning framework based on a decentralized communication topology. First, a distributed training method is proposed for neural networks over a decentralized communication topology based on the consensus algorithm, and it is proved that the distributed training allows all the agents connected in a decentralized communication topology to converge to the optimal model. Furthermore, the distributed training method is promoted based on the heuristic adaptive consensus algorithm, which allows the agent with better performance to have more influence on its neighbors. Moreover, the error-compensated model compression method with is applied in distributed training to compress the model parameter before sharing, which significantly saves communication costs with little decrease in model accuracy. The proposed distributed training framework with error-compensated communication compression is suitable for both IID and non-IID datasets. The distributed training strategy is also extended to reinforcement learning algorithms, and applied in some engineering applications, such as load forecasting, traffic light control, and autonomous vehicles.

  • 4:30-4:40pm: Recent advances in learning with graphs
    • Dr Xiaowen Dong, University of Oxford
      show more(on/off)

      Xiaowen Dong obtained the B. Eng. in Information Engineering from Zhejiang University, China, M.Sc. in Signal Processing & Communications from the University of Edinburgh, UK, and Ph.D. in Electrical Engineering from EPFL, Switzerland in 2000, 2008, and 2014 respectively. He was a Postdoctoral Associate, Media Lab, Massachusetts Institute of Technology, USA from 2014 to 2017, an Assistant Professor, Department of Engineering Science, University of Oxford, UK from 2017-2020, and currently an Associate Professor, Department of Engineering Science, University of Oxford, UK.

      Abstract: The effective representation, processing, analysis, and visualisation of large-scale structured data, especially those related to complex domains such as networks and graphs, are one of the key questions in modern machine learning. Graph signal processing (GSP), a vibrant branch of signal processing models and algorithms that aims at handling data supported on graphs, opens new paths of research to address this challenge, and has inspired the emerging field of graph neural networks (GNNs). In this talk, I will provide a high-level overview of GSP and GNNs, and discuss their applications in dealing with graph-structured data in real-world scenarios.

  • 4:40-4:50pm: Integrative network-based approaches to systems biology and systems medicine
    • Professor Huiru Zheng, Ulster University
      show more(on/off)

      Professor Huiru (Jane) Zheng is a Professor of Computer Science with the School of Computing, Ulster University. She received BEng in Biomedical Engineering and Instrumentation from Zhejiang University in 1989, MSc in Biomedical Information Processing from Fuzhou University in 1992 and was awarded a Ph.D. in Computer Science (Data Mining in Bioinformatics) in 2003 and a Postgraduate Certificate in Teaching in Higher Education in 2005 from Ulster University. Within her broad interests in machine learning, Prof. Zheng has particular research interests and expertise in integrative data analysis, complex network analysis, pattern recognition, and applications to biology, medicine, and healthcare. Her current research includes network analysis in metagenomics, agriculture, FinTech and similar disease, mHealth in gait analysis, mental health, and self-management of chronic diseases. She is currently leading the Data Analytics and Systems Theme in AI Research Centre at Ulster University.

      Abstract: Systems biology and systems medicine has merged as an interdisciplinary field, which promotes an integrative and holistic approach to studying biology and medicine at the systems level with the aim of improving our understanding, prevention, and treatment of complex diseases. Over the past years, we have developed and applied network-based approaches to integrative analysis of lifestyle data, multi-omics data (e.g. genomics, transcriptomics, proteomics, and metabolomics), and medical data to tackle complicated tasks such as the discovery of biomarkers, understanding of biological interactions, identification of complex disease patterns. This presentation will introduce a cluster of research projects undertaken by the group with a focus on multi-layer network approaches. Some results on metagenomics analysis, disease subtypes identification, and similarity disease identification will be presented. The presentation will conclude with a discussion of challenges and opportunities in multiscale computing and multiplex networks.

  • 4:50-5:00pm: Evolutionary computation for challenging optimization problems
    • Professor Shengxiang Yang, De Montfort University
      show more(on/off)

      Shengxiang Yang got his PhD degree from Northeastern University, China in 1999. He is now a Professor of Computational Intelligence (CI) and Director of the Centre for Computational Intelligence (CCI), School of Computer Science and Informatics, De Montfort University, UK. He has worked extensively for over 25 years in the areas of CI methods, including evolutionary computation and artificial neural networks, and their applications for real-world problems (e.g., production scheduling, communication network optimisation, and transportation network optimisation problems). His work has been supported by EPSRC, Royal Academy of Engineering, Royal Society, EU FP7 & Horizon 2020, and industry (e.g., BT, Honda, RSSB, and Network Rail), etc, with a total funding of over £2M. He has over 330 publications with an H-index of 57 according to Google Scholar.

      Professor Yang has served as an AE or Editorial Board Member for over ten international journals, including IEEE Transactions on Evolutionary Computation, IEEE Transactions on Cybernetics, Information Sciences, and Evolutionary Computation. He was the Founding Chair of the Task Force on Intelligent Network Systems (2012-2017) and chair of the Task Force on EC in Dynamic and Uncertain Environments (2011-2017) of the IEEE Computational Intelligence Society. He was the Founding Co-chair of IEEE Symposium on Computational Intelligence in Dynamic and Uncertain Environments (IEEE CIDUE), held respectively as part of the 20112021 IEEE Symposium Series on Computational Intelligence (IEEE SSCI). He has given over 20 invited keynote speeches and tutorials at international conferences.

      Abstract:Evolutionary computation (EC) is a class of stochastic optimization methods that are inspired by biological evolution. EC solves optimization problems by generating, evaluating and modifying a population of possible solutions. EC has been applied to solve different challenging optimization problems, e.g., Multi-objective Optimisation Problems (MOPs) where multiple conflicting objectives need to be optimized simultaneously and Dynamic Optimisation Problems (DOPs) where changes occur over time regarding the optimization objectives, decision variables, and/or constraint conditions, etc. These problems are challenging due to their nature of difficulty. Yet, they are important since researchers and practitioners in decision-making in many domains need to face them. This talk briefly overviews our research on EC for challenging optimization problems, including 1) EC for DOPs; 2) EC for MOPs; and 3) EC for Dynamic Multi-objective Optimisation Problems (DMOPs).

Day 2: Saturday 3rd July 2021 (Zoom link)

1:00-3:00pm: Keynote speeches, Chair: Professor Tao Cheng, University College London

  • 1:00-2:00pm:  Coevolutionary problem-solving
    • Professor Xin Yao, University of Birmingham and Southern University of Science and Technology (SUSTech), China
      show more on/off
      Professor Xin Yao
      Professor Xin Yao

      Xin Yao is a Chair Professor of Computer Science at the Southern University of Science and Technology (SUSTech), Shenzhen, China, and a part-time Professor of Computer Science at the University of Birmingham, UK. He is an IEEE Fellow and was a Distinguished Lecturer of the IEEE Computational Intelligence Society (CIS). He served as the President(2014-15) of IEEE CIS and the Editor-in-Chief (2003-08) of IEEE Transactions on Evolutionary Computation.

      His major research interests include evolutionary computation, ensemble learning, and their applications to software engineering. His research 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; and many other best paper awards at conferences. He received a Royal Society Wolfson Research Merit Award in 2012, the IEEE CIS Evolutionary Computation Pioneer Award in 2013 and the 2020 IEEE Frank Rosenblatt Award.

      Abstract:Coevolution is an old but very interesting research topic in evolutionary computation. This talk presents some of the applications of coevolution in learning and optimisation. First, we look at a classical coevolutionary learning scenario when no training data are available. In fact, no teacher information is available either. Then we examine how coevolution could be used to tackle large-scale global optimisation in the black box optimisation setting. Finally, we explore how coevolution could be harnessed to design general solvers automatically for hard combinatorial optimisation problems.

  • 2:00-3:00pm: Which is the tallest building in Europe? —Representing and Reasoning About Knowledge
    • Professor Ian Horrocks, University of Oxford
      show more on/off
      Professor Ian Horrocks
      Professor Ian Horrocks

      Ian Horrocks is a professor at the University of Oxford, and a fellow of Oriel College. He is also a Fellow of the Royal Society, a member of Academia Europaea, and a fellow of the European Association for Artificial Intelligence (EurAI). His research concerns the representation of knowledge, and the efficient manipulation of such knowledge by computers. He played a leading role in establishing the Semantic Web as a significant research field, pioneering many of the underlying logics, algorithms, optimization techniques, and reasoning systems.

      Abstract: the need for representing knowledge is ubiquitous in applications: travel apps need to know about transport links and accommodation; music apps need to know about artists and albums; (social) networking apps need to know about user profiles and preferences; and “virtual assistants” need to answer questions such as “what is the height of the Eiffel Tower”. Although real-world knowledge can be complex, a great deal can be achieved even with simple representations such as knowledge graphs; Google’s knowledge graph, for example, contains billions of facts, and enables Google to provide direct answers to many questions. Maintaining and extending such a knowledge graph is, however, extremely challenging. This problem can be mitigated to some extent by using structural rules, often called an ontology, to reduce the need for and identify obvious errors in explicitly stored facts. However, answering queries over ontology augmented knowledge graphs requires more complex reasoning, and scaling this up to large graphs is also challenging. In this talk I will introduce basic knowledge graphs, illustrate the benefits of augmenting them with ontologies, explore some inherent challenges, and compare various techniques for query answering over ontology augmented large-scale knowledge graphs.

3:00-3:10pm: Break

3:10-5:00pm: Short presentations, Chairs: Professor Hui Wang and Professor Huiyu Zhou

  • 3:10-3:20pm: Unleash more power of computational AI for creativity
    • Professor Yun Li, i4AI Ltd, U.K. / Shenzhen Institute for Advanced Study, UESTC, China
      show more(on/off)

      Yun Li was an intelligent systems Lecturer, Senior Lecturer and Professor with University of Glasgow from 1991 to 2018, and served as its Founding Director of joint education with UESTC in Chengdu and Founding Director of the University of Glasgow Singapore. Then, he served as Founding Director of Dongguan Industry 4.0 Artificial Intelligence Laboratory, China, and as Presidential Overseas Distinguished Professor at Dongguan University of Technology. He is currently Technical Director of i4AI Ltd, London, U.K., and Professor at Shenzhen Institute for Advanced Study, University of Electronic Science and Technology of China.

      During 2016-2019, Professor Li co-led with the Universities of Nottingham and Newcastle UK funding council’s “Connected Everything: Industrial Systems in the Digital Age Network+” national key program (£1m) to explore UK’s Industry 4.0 directions with AI, and on this role chaired the EPSRC’s “Looking Beyond Industry 4.0” conference, held at University of Glasgow in 2017. He is the author of the popular interactive evolutionary algorithm (EA) courseware EA_demo published online first in 1997 for demonstrating coded evolutionary algorithms and Computer-Automated Design (CAutoD). He has own 16 patents in 6 countries, and is the author or co-author/translator of 4 books and 260 papers, one of which is seen the most popular article in the IEEE Trans Control Systems Technology every month.

      Professor Li a Fellow of the Institute of Electrical and Electronic Engineers (IEEE) and currently serves as Associate Editor of IEEE Trans Neural Networks and Learning Systems, and of IEEE Tran Emerging Topics in Computational Intelligence. Email:

      Abstract: The progress of science and technology now relies heavily on the maturing simulation-based ‘computer-aided design’ (CAD), which can be augmented to ‘computerautomated design’ (CAutoD) via computational AI, so as to elevate design creativity and factoryfloor innovation in a manufacturing value chain. One opportunity identified of computational AI at the ‘Key Challenges and Future Directions of Evolutionary Computation’ Workshop that I organized for 2016 World Congress on Computational Intelligence was structural exploration to unleash more creativity, while one challenge identified was real-time adaptation. Concerning fundamental research into these issues, the talk will cover algorithm benchmarking, which plays a pivotal role in developing better techniques and in unleashing more power for real-world applications, as reflected by the recent call for the BENCH Special Issue of IEEE Transactions on Evolutionary Computation. The talk will deepen into comparison strategies and compatibility issues, illustrated with two possible paradoxes, namely, ‘cycle ranking’ and ‘survival of the nonfittest’. Further, a holistic set of performance metrics are presented for real-world applications, aiming at unleashing more power of computational AI paradigms for creativity.

  • 3:20-3:30pm: AI in research vs AI in reality – Insights from cases in non-tech industries
    • Dr Xiao Ma, University of Warwick
      show more(on/off)

      Dr Xiao Ma is an Associate Professor at WMG (formerly the Warwick Manufacturing Group), at the University of Warwick. His expertise is in digital economy, business transformation, innovation and entrepreneurship.

      He has been a bridge between industry and academia, by grounding practitioners’ knowledge and expertise into education and research programmes. Xiao is the head of the WMG Accelerator programme to fast-track growth for innovative businesses.

      He is a seasoned Entrepreneur, a 5-time founder to start, grow and exit ventures in UK and China. He also sits on the investment committee for Touch Stone Capital.

      Abstract: New ventures, usually technology companies, have leveraged Artificial Intelligence based solutions as a key differentiator when industry-hop into a non-tech industry. For example, Yingzi Tech entered big farming industry a few years back with a big facial recognition and auto-weighing promises. A group of top talents in AI from Silicon Valley raised the biggest angel investment for smart animal husbandry in China at the time, prior launching their products in 2018. However, such AI driven solutions are yet to be brought to reality. Industry has demanded a much higher accuracy for both tech solutions, i.e. 95% accuracy for weight through camera imagery, but such a high level has been proven to be challenging. The discrepancy for AI based cattle weighing is even worse. Widening the horizon, despite the advancement of applied AI research, many non-tech driven industries still find AI based solutions less adequate for implementation. These businesses range from smart agriculture, multi-modal mobility / transportation, smart machineries, bio-tech, etc.

      This talk will start with a few industry challenges that were expected to be addressed by AI, but experienced inadequate AI-based solutions. The talk will then electorate on the gaps between AI research and its applications in industry. And furthermore, discuss a few key raising opportunities for applied research.

  • 3:30-3:40pm: Contextual probability and neighbourhood counting
    • Professor Hui Wang, Ulster University
      show more(on/off)

      Hui Wang is Professor of Computer Science at Ulster University. His research interests are machine learning, knowledge representation and reasoning, combinatorial data analytics, and their applications in food authentication, virus detection and multimodal video search by examples. He has over 300 publications in these areas. He is principal investigator of a number of regional, national and international projects in the areas of image/video analytics (EPSRC funded MVSE 2021-2024, Horizon 2020 funded DESIREE and ASGARD, FP7 funded SAVASA, Royal Society funded VIAD), spectral data analytics (EPSRC funded VIPIRS 20202022), text analytics (INI funded DEEPFLOW, Royal Society funded BEACON), and intelligent content management (FP5 funded ICONS); and is co-investigator of several other EU funded projects. He is Head of the AI Research Centre in School of Computing, Ulster University. He is an associate editor of IEEE Transactions on Cybernetics. He is the Chair of IEEE SMCS Northern Ireland Chapter (2009-2018), and a member of IEEE SMCS Board of Governors (2010-2013).

      Abstract: In this talk I will present the concept of contextual probability, the resulting notion of neighbourhood counting and the various specialisations of this notion which result in new functions for measuring similarity. Contextual probability was originally proposed as an alternative way of uncertainty reasoning. It was later found to be an alternative way of estimating probability via neighbourhood counting, which is a kernel function and a generic similarity metric that can be applied to different types of data.

  • 3:40-3:50pm: AI for sound detection, tagging and classification
    • Professor Wenwu Wang, University of Surrey
      show more(on/off)
      Professor Wenwu Wang
      Professor Wenwu Wang

      Wenwu Wang is a Professor in Signal Processing and Machine Learning, and a Co-Director of the Machine Audition Lab within the Centre for Vision Speech and Signal Processing, University of Surrey, UK.

      He received the B.Sc. degree in 1997, the M.E. degree in 2000, and the Ph.D. degree in 2002, all from the College of Automation, Harbin Engineering University, China. He worked in King’s College London (2002-2003), Cardiff University (2004-2005), Tao Group Ltd. (now Antix Labs Ltd.) (2005-2006), Creative Labs (2006-2007), and University of Surrey (since May 2007). He was a Visiting Scholar at Ohio State University, USA, in 2008. His current research interests include blind signal processing, sparse signal processing, audio-visual signal processing, machine learning and perception, artificial intelligence, machine audition (listening), and statistical anomaly detection. He has (co)-authored over 250 publications in these areas.

      He is a (co-)recipient of over 15 awards including the Judge’s Award on DCASE 2020, the Reproducible System Award on DCASE 2019 and 2020, Best Student Paper Award on LVA/ICA 2018, the Best Oral Presentation on FSDM 2016, Best Student Paper Award finalists on ICASSP 2019 and LVA/ICA 2010, the TVB Europe Award for Best Achievement in Sound in 2016, and the Best Solution Award on the Dstl Challenge in 2012.

      He is a Senior Area Editor for IEEE Transactions on Signal Processing, an Associate Editor for IEEE/ACM Transactions on Audio Speech and Language Processing, an Associate Editor for EURASIP Journal on Audio Speech and Music Processing. He is a Specialty Editor in Chief for Frontiers in Signal Processing. He is an Elected Member of the IEEE Signal Processing Theory and Methods Technical Committee, IEEE Machine Learning for Signal Processing Technical Committee, and International Steering Committee of Latent Variable Analysis and Signal Separation. He was a Publication Co-Chair for ICASSP 2019, Brighton, UK, and a Satellite Workshop Co-Chair for INTERSPEECH 2022, Incheon, Korea.

      Abstract: Sound scene analysis, event detection and tagging have attracted increasing interest recently, with a variety of potential applications including situational awareness in defence, sound objects and events detection and identification in security surveillance, and acoustic sensing for smart homes and cities. This talk will provide an overview about some recent developments for several challenging problems related to this topic, including data challenges (e.g. DCASE challenges), acoustic modelling, feature learning, and dealing with weakly labelled data.

  • 3:50-4:00pm: Explore intersections between AI and control in both directions
    • Dr Tai Yang, University of Sussex
      show more(on/off)

      Dr Tai Yang has ten-year industrial experience. His long R+D career coves a wide range of engineering topics. As a distinguished innovator, he was invited to attend a Parliamentary Reception at the House of the Lords, West Minister. He was a member of a special renewable energy mission to visit Israel.

      Abstract: Based on a presenter’s recent paper “Four Generations of Control Theory Development”, the 4th generation of control is “Control in the New AI era”. This gives a control system view of the intersection between control and AI. On the other hand, the presenter will justify that Reinforcement Leaning (RL) is applying control theory to machine learning. Therefore standing on the AI side, among Unsupervised, Supervised and Reinforcement Learning, RL is the intersection between AI and control.

  • 4:00-4:10pm: Impacts of Covid-19 on human activity zones in London
    • Professor Tao Cheng, University College London
      show more(on/off)

      Tao Cheng is a Professor in GeoInformatics, the founder and Director of SpaceTimeLab for Big Data Analytics ( at University College London, a multi-disciplinary research centre that aims to gain actionable insights from geolocated and time-stamped data for government, business and society. Her research interests span AI and Big Data, network complexity, urban analytics (modelling, prediction, clustering, visualisation and simulation) with applications in transport, crime, health, business, social media, and natural hazards. She has studied and lectured in China, the Netherlands, Hong Kong, France and the UK. She has secured more than £20M research grants and published over 250 articles, working with many government and industrial partners in the UK including Transport for London, the London Metropolitan Police Service, Public Health England and Arup. Tongxin Chen is a PhD students in urban analytics in SpaceTimeLab.

      Abstract: Exploring the human activity zones (HAZs) gives significant insights into understanding the complex urban environment and reinforcing urban management and planning. Though previous studies have reported the significant human activity shifting at the city-level in global metropolises due to COVID-19 containment policies, the dynamic of human activity across urban areas at space and time during this ever-changing pandemic has not been fully examined. This study reveals the human activities zones from the mobile phone GPS trajectory data at different stages of the pandemic. The results show the HAZs not only exhibit declines in human activity but are strongly associated with urban land-use and population variables during the COVID-19 pandemic.

  • 4:10-4:20pm: Uncertainty quantification for trustworthy AI
    • Dr Yunpeng Li, University of Surrey
      show more(on/off)

      Yunpeng Li is a Lecturer in AI in the Department of Computer Science at the University of Surrey. His research interests are in the areas of statistical machine learning and signal processing, particularly Bayesian inference techniques and the optimal transport theory. He has broad interests in interdisciplinary applications of machine learning including disease detection, environmental sensing and object tracking. He received a PhD in Electrical Engineering at the McGill University in Canada in 2017 and was awarded the Junior Research Fellowship at the University of Oxford in 2018.

      Abstract: Uncertainty quantification constitutes a key element of trustworthiness in AI models. There are two main types of uncertainty that one can model – one is the model uncertainty given limited data; the other is the data uncertainty which captures noise inherent in data. We argue that a trustworthy AI system needs to measure the quality of uncertainty quantification, represent and propagate the model uncertainty, integrate the data uncertainty, and be validated through real-world applications. In this talk, I will introduce computationally efficient metrics for comparing high-dimensional distributions to facilitate uncertainty quantification, differentiable Bayesian filters to propagate model uncertainty, and Bayesian crowdsourcing to integrate data uncertainty. A real-world healthcare application will be presented to incorporate and validate the developed uncertainty quantification tools.

  • 4:20-4:30pm: Feature selection for visual object tracking
    • Dr Zhenhua Feng, University of Surrey
      show more(on/off)

      Dr Zhenhua Feng received his PhD degree in 2016, from the Centre for Vision, Speech and Signal Processing at the University of Surrey. He is currently a lecturer in Computer Vision and Machine Learning at the Department of Computer Science, University of Surrey. Before he joined the department, he was a Senior Research Fellow at the CVSSP, the University of Surrey. His research interests include biometrics, computer vision, pattern recognition and machine learning. He has published more than 60 original research papers in top-tier conferences and journals, including TPAMI, IJCV, CVPR, ICCV, IJCAI, ACL, TCYB, TIP, TIFS, TCSVT, TBIOM, ACM TOMM, Inf. Sci., Pattern Recognit., etc. He has received the 2017 European Biometrics Industry Award from the European Association for Biometrics (EAB) and the Best Paper Award for Commercial Applications from the AMDO2018 conference. He serves as editorial board members and reviewers for a number of top-tier conferences and journals in his research area, e.g., associate editor for the Springer journal, Complex & Intelligent Systems, area chair for BMVC2021, senior programme committee member for IJCAI2021.

      Abstract: Visual object tracking is one of the most popular research topics in computer vision. In this presentation, I present some recent studies in feature selection for high-performance visual object tracking. I first introduce an adaptive spatial feature selection method for tracking with discriminative correlation filters. Then an extended method that performs both spatial and channel selection will be presented for a further performance boost. Our approaches have achieved state-of-the-art results on various benchmarks. We also won the first place in the public dataset of the Visual Object Tracking (VOT2018) competition.

  • 4:30-4:40pm: Depression detection on Twitter – Knowing my mood from my words
    • Professor Huiyu Zhou, University of Leicester
      show more(on/off)

      Abstract: Depression is one of the most common mental illnesses in the world. Nearly 360 million people currently suffer from depression. In the United Kingdom, 7% of people meet the criteria of depression diagnosis, where 4-8% will experience depression in their lifetime. Depression sufferers who do not receive timely psychotherapy may develop severe conditions (e.g. insomnia and self-harm). Evidence shows that social networks (e.g. Twitter or Facebook) allow us to capture behavioural attributes such as mood, communication and socialisation. Many research studies for online depression detection have mainly focused on exploring attributes of depression behaviours but ignoring the fitness of classification tasks. In this talk, we present a novel classification algorithm based on Adaboost that can mitigate the influence of noise or errors and have a strong fitness and generalisation ability. Our proposed method is based on a cost-sensitive boosting pruning trees algorithm to effectively classify non-depressed and depression Twitter users.

  • 4:40-4:50pm: Data-driven evolutionary optimization – Integrating machine learning, evolutionary computation and data science
    • Professor Yaochu Jin, University of Surrey
      show more(on/off)
      Professor Yaochu Jin
      Professor Yaochu Jin

      Yaochu Jin received the B.Sc., M.Sc., and Ph.D. degrees from Zhejiang University, Hangzhou, China, in 1988, 1991, and 1996, respectively, and the Dr.-Ing. degree from Ruhr University Bochum, Germany, in 2001.

      He is currently a Distinguished Chair, Professor in Computational Intelligence, Department of Computer Science, University of Surrey, Guildford, U.K., where he heads the Nature Inspired Computing and Engineering Group and Director of Research. He was a “Finland Distinguished Professor” of University of Jyvaskyla, Finland, a “Changjiang Distinguished Visiting Professor”, Northeastern University, China, and “Distinguished Visiting Scholar”, University of Technology Sydney, Australia. In 2021, he was awarded the “Alexander von Humboldt Professorship for Artificial Intelligence” by the Federal Ministry of Education and Research, Germany.

      Professor Jin is presently the Editor-in-Chief of the IEEE TRANSACTIONS ON COGNITIVE AND DEVELOPMENTAL SYSTEMS and the Editor-in-Chief of Complex & Intelligent Systems. He was an IEEE Distinguished Lecturer in 2013-2015 and 2017-2019, the Vice President for Technical Activities of the IEEE Computational Intelligence Society (2015-2016). He was the General Co-Chair of the 2016 IEEE Symposium Series on Computational Intelligence and the Chair of the 2020 IEEE Congress on Evolutionary Computation. He is the recipient of the 2018 and 2021 IEEE Transactions on Evolutionary Computation Outstanding Paper Award, the 2015, 2017, and 2020 IEEE Computational Intelligence Magazine Outstanding Paper Award, and the Best Paper Award of the 2010 IEEE Symposium on Computational Intelligence in Bioinformatics and Computational Biology. He was named by the Web of Science as “a Highly Cited Researcher” in 2019 and 2020. He is a Member of Academia Europaea and Fellow of IEEE.

      Abstract: Data-driven optimization problems are commonly seen in the real-world, ranging from engineering design, to drug discovery, to automated architecture search of deep neural networks. This talk begins with a brief introduction to the fundamentals of data-driven optimization, followed by an overview of recent advances that exemplify how machine learning techniques, Bayesian optimization and evolutionary algorithms can be integrated to solve complex optimization problems. Finally, we show our most recent work on federated data-driven evolutionary optimization when the data are distributed and subject to privacy constraints.

4:50-5:00pm: Closing remarks

Organising Committee

Professor Yaochu Jin, University of Surrey (Chair)
Professor Tao Cheng, University College London
Professor Zhengtao Ding, University of Manchester
Professor Henry Li, University of Manchester
Professor Hui Wang, Ulster University
Professor Shengxiang Yang, University of De Montfort
Professor Huiru Zheng. Ulster University
Professor Huiyu Zhou, University of Leicester