Invited Speaker 1: Ming WU, Engineering Manager, Intel Corporation

Title: A virtual computing and storage approach to computing continuum

Email Addresses: ming.m.wu@intel.com

Abstract: The proliferation of information technology and global-wide deployment of broadband and 5G infrastructure make computer hardware unprecedented cheap and internet access ever-increasing easy. These changes have brought an imperceptible trend that more users might have multiple computing devices together accessing same copy of personal software environment from multiple devices at geographically different locations. Current technologies to address this request have inevitable limitations in network dependency, native computing experience or only for specific application only. In this session, an end-to-end virtual computing and storage (VCS) approach to cloud-edge-client collaborative computing continuum is proposed under the inspiration of transparent computing theory. It’s based on virtual computing at client side to make full use of hardware for entire software stack but keeps software and personalized setting of each individual user with virtual storage mechanism. With the collaboration between cloud-edge-client, the user-based desktop environment – the entire software stack - could migrate among multiple clients and even to heterogeneous clients within system wide network. We also provides optimizations, system evaluation and analysis to this approach.

Short-Bio: Ming Wu received his B.Sc. and M.Sc degrees in computer science, from Tsinghua University, China, in 1997 and 1999 respectively. He also held a minor degree of B.Economics from Tsinghua University in 1997. Ming joined Intel in 2004 and served as different technical leadership roles like senior engineer and engineering manager in SSG. He led the development of several UEFI embedded projects including Intel® Boot Loader Development Kit and Intel® Software Solution for Transparent Computing, and owned the collaboration with universities and enterprises in Transparent Computing. His working interests include computer architecture, computer network, distributed system and embedded systems, and he published several technical papers in these areas. Before joining Intel he worked for two startup companies to build internet router and network security equipment.

Invited Speaker 2: Fujin Huang, Intel Corporation

Title: An Innovative Way to Provide Value-add Services at Pre-boot Stage

Email Addresses: fujin.huang@intel.com

Abstract: There are increasing and strong requests to provide value-add services during pre-boot stage on modern platforms, such as facial recognition service for user authentication before booting to target OS, and security service for validating the integrity of OS image before booting. Most of these value-added services are quite complex software applications, which may need multitasking software architecture and high-performance peripherals such as AI acceleration cards and high-resolution camera. Also, it is required to loading the target OS directly without reboot after running value-add services. A typical idea to address these requests is to implement the services as software modules in BIOS level. However, BIOS is quite a low-level and enclosed system which looks like a black box to most developers and customers. It’s single threaded and most of its hardware drivers could not provide high performance experience. We here present a generic and scalable mechanism to provide high performance value-add services with multi-threading complex tasks during pre-boot stage.

Short-Bio: Fujin Huang received his B.Sc. degree in Control Science and Engineering (CSE) from Nanchang University, China, and his M.Sc. degree in CSE from Shanghai Jiaotong University, China, in 2005 and 2008, respectively. Fujin is the senior platform architect of Intel China Asia Pacific R&D Ltd. Since Fujin joined Intel in 2008, he contributed in various product projects from UEFI BIOS to Turn-Key solutions, and filed several technical patents. His research interests include cloud computing, big data and edge computing.

Invited Speaker 3: Hai Shen, Marketing Manager, Intel

Title: Intel® Ultra Cloud Cllent Solution and Its Industry Application

Email Addresses:

Abstract: Cloud client has huge market in multiple industries and its computing architectures and solutions keep evolving to be more effective, secure and robust. This session will talk about typical cloud client market status and industry application, introduce Intel® Ultra Cloud Client (UCC) Solution and its key competitive advantages, and shares some UCC key application use cases in education, healthcare, and banking market segments.

Short-Bio: Hai Shen is marketing manager in Internet of Things Group in Intel. He is now responsible for education market analysis and solution development, promotion and scaling with ecosystem partners. Hai graduated from Northwestern Polytechnical University in 2006 and holds a master degree in software engineering. Hai joined Intel in 2004 and have worked as training manager, software community manager, academic program manager, technical marketing engineer in various teams in Intel.

Invited Speaker 4: Prof. Sancheng Peng, Guangdong University of Foreign Studies

Title: Deep Broad Learning for Cross-Domain Emotion Classification

Email Addresses:

Abstract: Cross-domain emotion classification aims to leverage useful information in a source domain to help predict emotion polarity in a target domain in a unsupervised or semi-supervised manner. In this talk, we provide an overview on the different methods for cross-domain emotion classification. Then, we will discuss how to combine the advantages of deep learning and broad learning to conduct single-source and multi-source cross-domain emotion classification, respectively. Finally, we will provide the analysis for experimental results.

Short-Bio: Sancheng Peng received the PhD degree in computer science in 2010 from Central South University, China. He is a Professor of Guangdong University of Foreign Studies, China. He was a Research Associate of City University of Hong Kong from 2008 to 2009. He has authored or coauthored over 60 technical papers in both journals and conferences, such as the IEEE Communications Surveys and Tutorials, IEEE Transactions on Knowledge and Data Engineering, IEEE Transactions on Dependable and Secure Computing, IEEE Wireless Communications, IEEE Network, IEEE Internet of Things Journal, Journal of Network and Computer Applications, Computer Networks, Computer and Security, Information Sciences, Future Generation Computer Sciences, Journal of Computer and System Sciences, Journal of Computer Science and Technology, IEEE TrustCom, IEEE CBD, ICA3PP, SpaCCS, and EUC. His research interests include network and information security, natural language processing, social networks, and mobile computing. Dr. Peng has served as the Guest Editor of Future Generation Computer Systems and as a PC member for various prestige international conferences. He is a Senior Member of the CCF and a member of ACM.

Invited Speaker 5: Prof. Tian Wang, Beijing Normal University

Title: Sensor-Cloud and Edge Computing: Overview, Solutions, and Directions

Email Addresses:

Abstract: Sensor-cloud originates from the extensive use of Wireless Sensor Networks (WSNs) and cloud computing. However, there are obvious limitations in both WSNs (e.g., communication, storage, energy, computation, and scalability) and cloud computing (e.g., high delay, longdistance transmission, and privacy disclosure). In this talk, Prof. Wang examines the origins of sensor-cloud and provides an in-depth and comprehensive discussion of three key challenges, namely reliability, energy, and heterogeneity. He also introduces some initial designs of new edge-based schemes to address these challenges and concludes the talk with a discussion on the remaining challenges and future research directions.

Short-Bio: Prof. Wang received his BEng and MEng degrees in computer science and technology from Central South University and his Ph.D. degree in Computer Science from the City University of Hong Kong. He is currently a full professor at Beijing Normal University. Prof. Wang is the top 2% scientist according to "World's Top 2% Scientists 2021," published by Stanford University. He was supported by the "Hundred-Thousand-Ten Talent Project" and the Science Fund for Distinguished Young Scholars of Fujian Province. His research covers a wide range of topics, including the Internet of Things, Edge Computing, Mobile Computing. He has published over 200 papers in reputed high-level journals and conferences, including 30 IEEE/ACM Transactions papers. He has more than 8000 citations (H-Index is 49), according to Google Scholar.

Invited Speaker 6: Dr. Wenjun Jiang, Hunan University

Title: Recommendation Technique for Social Good: Reflections and Practice

Email Addresses:

Abstract: Personalized recommendation is the key technology to solve the information overload of online system and improve the online user experience. However, existing works usually focus on improving the recommendation accuracy, leading to several issues such as the “Filter Bubble”, “Information Cocoon”, “Echo Chamber”, as well as the decrease of satisfaction and the long tail effect of the platform. Then, we need to pay more attention to improve recommendation and exploit it to promote user growth and social development. In this talk, I will share some reflections on “Recommendation Technique for Social Good.” I will also share some of our recent works on that, including serendipity recommendation, recommendation for online learning, and product review analysis for recommendation (e.g., spam detection and helpfulness evaluation).

Short-Bio: Wenjun Jiang received her Bachelor’s degree in Computer Science from Hunan University, P. R. China, in 2004, Master’s degree in Computer Software and Theory from Huazhong University of Science and Technology, P. R. China, in 2007, and Doctor’s degree in Computer Software and Theory from Central South University, P. R. China, in 2014. She has been a visiting Ph. D student at Temple University for two years. After graduation, she was an assistant professor in Hunan University. Since January 2017, she is an associate professor and a Doctoral Supervisor in Hunan University. Her research interests include social network analysis, recommendation systems, and smart education and learning optimization. She has published more than 40 technical papers in the above areas, including top international journals like ACM CSUR, IEEE TC, IEEE TPDS, ACM TKDD, ACM TOIT, ACM TWeb and top international conferences like WWW, INFOCOM, AAAI, CIKM. Her research is supported by Key Project of the National Natural Science Foundation of China, National Natural Science Foundation of China, National Natural Science Foundation of Hunan Province, Open project of Zhejiang Lab, and Science and technology program of Changsha city. Dr. Jiang is a Senior Member of CCF, and Member of the IEEE and ACM.

Invited Speaker 7: Dr. Yinglong Dai, Hunan Narmal University

Title: Hierarchical Deep Reinforcement Learning with State and Action

Email Addresses:

Abstract: Deep reinforcement learning (DRL) is a powerful tool to solve the problems of high-dimensional data perception and complex dynamic decision-making. However, DRL methods still face challenges when the combinatorial state-action space becomes excessively large, such as real-world environments and non-stationary environments. Hierarchical reinforcement learning (HRL) is an effective approach to resolve the scalability challenges by decomposing a complicated task into relatively simple tasks. How to build the hierarchical structure of an agent's decision-making process becomes a main problem of hierarchical DRL methods. To improve the training efficiency and the model interpretability, we propose to use the conceptual embedding techniques to build the hierarchical structure, in which we introduce prior knowledge explicitly and restrict the exploration space reasonably. We split the DRL policy into two main functional modules. One is the recognition module that is used to recognize the latent state of the environment by the high-dimensional observation data. The state representation space of the recognition module forms a hierarchy by aggregating similar observation features to different levels. Another is the decision module that is used to plan action strategies according to the latent state of the environment. The decision module forms a hierarchy by decomposing the ultimate goal into sub-goals. In this way, the policy of the DRL agent would have a clear inference pipeline, in which we can incorporate prior knowledge into the deep model and improve the model interpretability.

Short-Bio: Yinglong Dai received B.S. and M.S. degrees in automation and control theory & control engineering from Northeastern University, China, in 2010 and 2012, respectively. He received a Ph.D. degree in computer science from Central South University, China, in 2018. From 2012 to 2013, he was an Electronic Engineer with the Research Institute of Intelligent Engineering, Sany Heavy Industry, Changsha, China. At present, he is a lecturer with College of Information Science and Engineering, Hunan Normal University, Changsha, China. His research interests include multimodal deep learning, deep reinforcement learning, healthcare, and multi-agent systems. At present, he has published over ten SCI/EI papers
(https://www.researchgate.net/profile/Yinglong_Dai).

Invited Speaker 8: Dr. Qin Liu, Hunan University

Title: Secure Search in Cloud Computing

Email Addresses:

Abstract: Cloud computing, providing a wide variety of services in a pay-as-you-go fashion, is an extremely successful paradigm of service-oriented computing. With the increasing popularity of cloud-based services, consumers are highly motivated to outsource their data and computing services to cloud platforms. To protect user privacy from the cloud service provider (CSP), existing research suggests encrypting data before outsourcing. This makes traditional data services like keyword-based searches very challenging. The simple solution of downloading all the encrypted data and decrypting them locally is extremely expensive. Therefore, investigating an efficient search service over ciphertexts becomes a paramount urgency. This talk will investigate security and privacy issues in cloud computing, and attempt to identify possible solutions to achieve secure search services in cloud computing.

Short-Bio: Qin Liu received her B.S. in Computer Science in 2004 from Hunan Normal University, China, received her M.S. in Computer Science in 2007, and received her Ph.D. in Computer Science in 2012 from Central South University, China. She has been a Visiting Student at Temple University, USA. Now, she is an Associate Professor in the College of Computer Science and Electronic Engineering at Hunan University, China. Her research interests include security and privacy issues in cloud computing and social networks, and big data security. She has published more than 60 technical papers and books/chapters in the above areas, including top international journals and conferences like IEEE TPDS, IEEE TSC, ACM CCS, IEEE INFOCOM, and so on. She has been serving as a Guest Editor, Conference vice Co-chair, Workshop Co-Chair, Publicity Chair/Co-Chiar, TPC, and reviewer for international journal/conference proceedings.

Invited Speaker 9: Prof. Entao Luo, Hunan University of Science and Engineering

Title: Research on Key Technologies of Dynamic Modeling for User Privacy in Mobile Application Services

Email Addresses:

Abstract: Traditional privacy protection schemes for mobile users often only focus on protecting the privacy of static data sets at a certain stage, but it is difficult to dynamically measure and adaptively protect the dynamic evolution of private data during the whole process. This research t intends to draw on the latest research results in the fields of deep learning, knowledge graphs, data mining, data publishing, etc., and commits to establishing a set of key technology systems suitable for the dynamic protection of mobile users’privacy. First, a dynamic measurement model based on deep learning is proposed to enhance the protection of user privacy and the selfadaptive evolution characteristics of the model. Secondly, the protection mechanism of multidimensional sensitive data generalized hierarchical tree is proposed to realize the user privacy protection requirements with non-relevant and non-differentiation. next, a data mining scheme based on decision tree under differential privacy is proposed, which can realize effective data value mining without exposing user privacy. Finally, a discontinuous data dynamic publishing scheme with hidden user relationship graphs is proposed to prevent attackers from guessing the relationship of user data and improve publishing efficiency.

Short-Bio: Entao Luo, He is a professor at the School of Information Engineering, Hunan University of Science and Technology, He received his Doctor’s degree in Software Engineering from Central South University, P. R. China, His research fields include mobile social networking, machine learning, and edge computing security and privacy preserving. He has published more than 30 technical papers in the above areas, including Journal of Software、Journal of Computer Research and Development、Information Science、FGCS、IEEE Communications Letters etc. In recent years,His research is supported by Key Project of the National Natural Science Foundation of China, National Natural Science Foundation of Hunan Province, Open project of Guangxi Key Cryptography Laboratory, and Science and technology program of Yongzhou city. Dr. Luo is a Member of CCF, and Member of the IEEE and IEEE VTS.

Invited Speaker 10: Dr. Shaobo Zhang, Hunan University of Science and Technology

Title: Research on Multi-Dimensional User Privacy Protection for Large-Scale Mobile Social Networks on Edge Computing

Email Addresses:

Abstract: The user's privacy risks and query efficiency issues that exist in the mobile social network query process have received extensive attention from the academic community. Due to the large scale, diversity and high speed of large-scale mobile social networks, traditional user privacy protection schemes lack dynamic privacy and only have a single protection goal, which is difficult to support the personalized privacy protection and efficient query needs of large-scale mobile social network users. The speaker first considers the increasingly complex cloud computing environment and massive dynamic big data as new challenges to user privacy query services based on large-scale mobile social networks, based on the existing cloud computing, edge computing model is introduced, and it performs hierarchical processing, storage and protection of large-scale mobile social network user data. Then, according to the personalized and differentiated privacy protection needs of users in different query processes, from the user location privacy, query privacy and attribute privacy three dimensions to protect user privacy, build a multidimensional privacy protection system for users of large-scale mobile social networks oriented to edge computing, so as to provide secure, efficient and flexible query services for large-scale mobile social network users in the cloud-side collaborative environment.

Short-Bio: Shaobo Zhang received the B.S. and M.S. degree in Computer Science both from University of Science and Technology, Xiangtan, China, in 2003 and 2009 respectively, and received the Ph.D. in Computer Science in 2017 from Central South University, China. Now, he is an Associate Professor in the School of Computer Science and Engineering at Hunan University of Science and Technology, China. His research interests include security and privacy issues in cloud computing, social networks and big data. He has published more than 60 papers in IEEE IOT Journal, Information Sciences, FGCS and other journals and international conferences. Dr. Zhang is a Member of the CCF.

Invited Speaker 11: Dr. Qiang Zhang, Nanchang Municipality

Title: The Development Trends and Example of Smart Cities

Email Addresses:

Abstract: A smart city focuses on innovation and integrates information and communication technologies into a comprehensive development strategy geared toward transforming urban development. Additionally, the new smart city promotes the happiness and satisfaction of its inhabitants in order to promote a new path and new model for sustainable urbanization. The National Fourteenth FiveYear Plan proposes to "accelerate digital development and build a digital China" and "enable more people to enjoy a higher-quality urban life" and requires "building smart cities and digital villages" and "overall improvement of urban quality" and to provide guidelines for the development of a new type of smart city in the new era. As a starting point for this report, let's look at the development trends in smart cities, highlight the hot spots of research in smart cities, using the phased results of the smart city construction in Nanchang as an example to intuitively illustrate the current development trend of smart cities.

Short-Bio: Qiang Zhang received the B.S. degree from Central South University, China in 2011, and the M.S. degree from University of Chinese Academy of Sciences, China, in 2014. and the Ph.D degree from Central South University, China, in 2019. Now, he is a section chief of Big Data Development Administration of Nanchang Municipality. He was a visiting Ph.D. student in the Department of Electrical and Computer Engineering, University of Waterloo, Canada. His research interests include privacy preserving, cloud computing and information retrieval.

Invited Speaker 12: Dr. Feng Wang, China University of Geosciences

Title: sEMG-based motion intent recognition methods in non-ideal conditions

Email Addresses:

Abstract: Motion intent recognition is a key technology for intelligent rehabilitation robots. In sEMG (surface electromyography)-based recognition, most studies aim at improving recognition accuracies. While in real applications, sEMG-based recognition systems are limited by many disturbances in non-ideal conditions. We need to focus on the robustness of sEMG-based recognition. In this talk, I will share some disturbances in non-ideal conditions, including electrode shifts, individual differences, muscle fatigue, limb postures and others. I will also share many novel methods that are proposed to remove or reduce the impact of these disturbances, including building sEMG-based datasets, exploiting deep-learning-based and transfer- learning-based recognition, and sEMG decomposition.

Short-Bio: Feng Wang received his Ph.D. degree in computer science from Central South University, Changsha, China, in 2018. He was an overseas researcher (JSPS Fellow) with the School of Engineering, Tokyo University of Technology, Japan, from 2019 to 2021. He is currently an associate professor with the School of Automation, China University of Geosciences, Wuhan, China. His research interests include social network analysis, machine learning, Internet of Things, motion intent recognition and recommendation systems. He has published more than 20 technical papers in the above areas, including top international journals like Information Sciences and top international conferences like IFAC World Congress. His research is supported by National Natural Science Foundation of China, China Postdoctoral Science Foundation, Japan Grantsin-Aid for JSPS Research Fellows. Dr. Wang is a Member of IEEE Industrial Electronics Society, and Member of the IEEE.

Invited Speaker 13: Dr. Guihua Duan, Central South University

Title: New application of classical cryptography in privacy protection

Email Addresses:

Abstract: In this talk, we firstly introduce the basic cryptographic protocols, and then discuss the deformation and application of advanced cryptographic protocols such as blind signature protocol, inadvertent transmission protocol and secure computing protocol.

Short-Bio: Guihua Duan received her PhD in Computer Science and Technology from Central South University in 2010. She is currently an Associate Professor in the School of Computer Science and Engineering, Central South University. Her research interests include security and privacy, and big data. She has published more than 20 academic papers, presided over or participated in 4 National Natural Science Foundation projects, served as guest editor of IJAACS special issue and program co-chair of TrustCom2012, and won the first prize of provincial teaching competition and the title of teaching expert of teachers in colleges and universities in Hunan Province.

Invited Speaker 14: Dr. Xiaolong Hu, Central South University

Title: Virtualization of Block Device in Edge AI Computing

Email Addresses:

Abstract: Edge computing devices have strong heterogeneity and different architecture. It is a great challenge to deploy reconfigurable operating systems and applications on massive edge devices. In addition, with the development of operating system, the high performance and consistency of storage and file system software stack is also a hot innovation field. This talk introduces the origin and essence of block device virtualization technology, discusses the use of block device virtualization / transparent computing technology to solve the problems faced in the deployment of edge computing devices. An edge image AI computing device based on hardened virtual block equipment is designed and implemented for intelligent identification of EHV GIS(Gas Insulation Switch) contact terminal.

Short-Bio: Xiaolong Hu received her Bachelor’s degree in Computer Science and Engineering from Harbin Institute of Technology, Master’s degree and Doctor’s degree in Space Physics from University of Science and Technology of China. He is now an associate professor in Central South University. His research interests include Operating System, Embedded system, Image processing and machine learning. He has published more than 50 technical papers in space physics and computer science.

Invited Speaker 15: Dr. Lei Zhou, Southern University of Science and Technology

Title: Study on Hardware-assisted Trusted Execution Environment in x86 Platform

Email Addresses:

Abstract: Trusted Execution Environments (TEEs) have been widely adopted in commodity systems for enhancing the security of software execution. The examples of TEE technologies from COTS in x86 platform include but not limited to: Intel Software Guard eXtensions (SGX), AMD Memory Encryption Technologies, x86 System Management Mode, AMD Platform Secure Processor, and Intel Management Engine (ME). However, the users are difficult to access the TEE directly due to the isolation and close source. We study those existing hardware features in the x86 Platform (CPU, chipset), learn the debugging approaches to understand those technologies even lack of datasheets. Finally, we can leverage those to construct the hardware-assisted trusted execution environment, introducing a minimal TCB and incurring negligible overhead on the host system.

Short-Bio: Lei Zhou received the PhD degree in Computer Science from Central South University. He is a Post-doctoral Fellow in the Department of Computer Science and Engineering at Southern University of Science and Technology (SUSTech). He has been a Visiting Student at Wayne State University, USA. His primary research interests are in the areas of x86 systems security, including trustworthy execution, hardware-assisted security, and memory forensics.

Invited Speaker 16: Yang Xu, Hunan University

Title: Blockchain-empowered secure data sharing and privacy leakage accountability mechanism

Email Addresses:

Abstract: Data is an extremely important asset of the Internet ecosystem. As one of the basic services of the Internet, data sharing is widely used to help users make full use of data resources and avoid repeated collection. However, shared data faces the tampering risk in this open environment. Due to malicious attacks or sharer’s unintentional mistakes, the data obtained by consumers may be inconsistent with their expectations. Besides, the illegal dissemination of shared data, such as information leaks and piracy, also discourages providers to share their data. Tracking down the sources of leaks in an open network becomes quite a daunting task. The existing leakage traceability schemes are difficult to use due to the lack of Trusted Third Party (TTP) and unaffordable overhead in practice. Fortunately, the emergence of blockchains has made it possible to mitigate these problems. In this talk, we introduce a blockchain-empowered secure data sharing and privacy leakage accountability mechanism which has been applied to digital content sharing scenarios. In our approach, similarity hashing and asymmetric fingerprinting are used to ensure the correctness and traceability of data. Blockchain is utilized as a self-recording channel for achieving non-repudiation evidence of service interactions and a distributed content verifier. We also design a blockchain-based interactive protocol to achieve secure data sharing. According to these on-chain records, smart contracts can automatically trace the leakers and carry out corresponding punishments. We believe our mechanism can provide comprehensive protection for both content providers and consumers.

Short-Bio: Yang Xu received the Ph.D. degree in Computer Science and Technology from Central South University, China. From 2012 to 2013, he was a Software Engineer in Intel Cooperation (Asia Pacific R&D Center). From 2015 to 2017, he was a Visiting Scholar in the Department of Computer Science and Engineering at Texas A&M University, USA. He is currently an Associate Professor and Deputy Director of the Department of Cyberspace Security at the College of Computer Science and Electronic Engineering, Hunan University, China. His research interests include, cloud computing, blockchain, artificial intelligence, and privacy computing. His research is supported by the National Natural Science Foundation of China, the Natural Science Foundation of Hunan Province, etc. He has published over 50 articles in international journals and conferences, including IEEE IoTJ, TSC, TII, TCC, TETC, TCBB, TNSE, etc. He was the awardee of the Best Paper Award of IEEE International Conference on Internet of People (IoP 2018). He serves/has served as a Program Committee Chair for UbiSec 2021 and IWCSS 2020, a Track Chair for IEEE CyberSciTech 2020, the Publicity Chair for CPSCom 2020, Blocksys 2020, ISSR 2019, Ubisafe 2019, and a reviewer for over 20 international journal/conference proceedings. He is a member of Blockchain Technical Committee of China Computer Federation (CCF) and China Society for Industrial and Applied Mathematics (CSIAM), and a member of IEEE and ACM.