Speakers and Speeches Information

Bo An

School of Computer Science and Engineering, Nanyang Technological University

Title: When AI Meets Game Theory

Abstract: Artificial Intelligence is changing the world now! In January 2017 CMU’s Libratus system beat a team of four top-10 headsup no-limit specialist professionals, which was the first time an AI had beaten top human players in this game. Libratus’s success is purely based on algorithms for solving large scale games and has nothing to do with deep learning! Over the last few years, algorithms for solving large scale games have also been applied to scheduling security resources in many significant domains including airports, federal air marshals service, coast guard, and wildlife conservation organizations. Game theory has also been applied in many other AI problems such as sustainability, ad-word auction, and e-commerce. This talk will discuss key techniques behind these success and their potential applications in other domains.

Bio: Bo An is an Associate Professor with the School of Computer Science and Engineering, Nanyang Technological University, Singapore. He received the Ph.D degree in Computer Science from the University of Massachusetts, Amherst. His current research interests include artificial intelligence, multiagent systems, computational game theory , reinforcement learning, and optimization. His research results have been successfully applied to many domains including infrastructure security and e-commerce. He has published over 90 referred papers at AAMAS, IJCAI, AAAI, ICAPS, KDD, WWW, JAAMAS, AIJ and ACM/IEEE Transactions. Dr. An was the recipient of the 2010 IFAAMAS Victor Lesser Distinguished Dissertation Award, an Operational Excellence Award from the Commander, First Coast Guard District of the United States, the 2012 INFORMS Daniel H. Wagner Prize for Excellence in Operations Research Practice, and 2018 Nanyang Research Award (Young Investigator). His publications won the Best Innovative Application Paper Award at AAMAS’12 and the Innovative Application Award at IAAI’16. He was invited to give Early Career Spotlight talk at IJCAI’17. He led the team HogRider which won the 2017 Microsoft Collaborative AI Challenge. He was named to IEEE Intelligent Systems' "AI's 10 to Watch" list for 2018. He was invited to be an Advisory Committee member of IJCAI’18. He is a member of the editorial board of JAIR and the Associate Editor of JAAMAS, IEEE Intelligent Systems, and ACM TIST. He was elected to the board of directors of IFAAMAS and senior member of AAAI.

Xiang Bai

Department of Electronics and Information Engineering, Huazhong University of Science and Technology

Title: OCR in the Wild: Current Situation, Challenges and Future Trends

Abstract: Reading text in the wild, consisting of two main steps: scene text detection and scene text recognition,is a general OCR technology that attracts wide attention from academia and industry. Recently, remarkable progresses have been achieved for scene text reading due to the successes of deep neural networks. In this talk, I will give a thorough overview of the state-of-the-art deep learning methods for scene text reading, and summarize its urgent challenges in real applications. Last, the future trends of this area will be predicted.

Bio: Xiang Bai is currently a Full Professor with the school of Electronic Information and Communications, Huazhong Univerisity of Science and Technology (HUST), Wuhan, China. He received the BS, MS, PhD degree from HUST in 2003, 2005, 2009, respectively. In recent years, he has focused on scene text reading, and developed a series of state-of-the-art methods on text detection, text recognition, script identification in natural images. He serves as an associate editor for Pattern Recognition, Pattern Recognition Letters, and Frontier of Computer Science.

Enhong Chen

School of Computer Science and Technology, University of Science and Technology of China

Title: Exploiting Knowledge Tracing for Intelligent Education

Abstract: Intelligent education systems can help the personalized learning of students with computer-assisted technology. Along this line, it is well known that modeling the cognitive structure including the knowledge level of learners and the knowledge structure (e.g., the prerequisite relations) of learning items is important for intelligent education services (e.g., learning path recommendation). In this talk, I will first introduce our attempts to extract knowledge from learning items and show the way to trace the evolving knowledge levels of learners at each learning step. Then, I will give a general framework to fully exploit the multifaceted cognitive structure for learning path recommendation. Extensive experimental results demonstrate both the effectiveness and the superior interpretability of our proposed solutions.

Bio: I am a Professor and vice dean of School of Computer Science of University of Science and Technology of China(USTC), CCF Fellow, IEEE Senior Member (Since 2007), winner of the National Science Fund for Distinguished Young Scholars (in 2013), scientific and technological innovation leading talent of 'Ten Thousand Talent Program'(in 2017) and member of the Decision Advisory Committee of Shanghai (Since June, 2018). I am also the vice director of the National Engineer Laboratory for Speech and Language Information Processing, the director of Anhui Province Key Laboratory of Big Data Analysis and Application, and the chairman of Anhui Province Big Data Industry Alliance. I received my B.Sc degree from Anhui University in 1989, Master degree from Hefei University of Technology in 1992 and Ph.D degree in computer science from USTC in 1996. My current research interests are data mining and machine learning, especially social network analysis and recommender systems. I have published more than 200 papers on many journals and conferences, including international journals such as IEEE Trans, ACM Trans, and important data mining conferences, such as KDD, ICDM, NIPS. My research is supported by the National Natural Science Foundation of China, National High Technology Research and Development Program 863 of China, etc. I won the Best Application Paper Award on KDD2008 and Best Research Paper Award on ICDM2011.

Xudong Chen

Department of Electrical and Computer Engineering, National University of Singapore

Title: Solving Electrical Impedance Tomography via Deep Learning Approach

Abstract: Electrical Impedance Tomography (EIT) is a tomographic imaging method that numerically reconstructs the conductivity of an object using the boundary voltage-current data collected from the surface. EIT has wide and important applications, such as in medical imaging, material engineering, civil engineering, biotechnology, and chemical engineering. EIT is well known to be a challenging problem, due to severe ill-posedness and nonlinearity. This talk proposes the convolution neural network (CNN) technique to solve EIT problems for medical applications. In order to make machine learning more powerful, a deep understanding of the corresponding forward problem is desirable. In solving ISP, the concept of contrast current plays an essential role in the proposed CNN technique, which enables us to design the architecture of learning machine such that unnecessary computational effort spent in learning wave physics is minimized or avoided. Numerical simulations and experimental data demonstrate that the proposed CNN scheme outperforms a brute-force application of CNN. The proposed deep learning inversion scheme is promising in providing quantitative images in real time.

Bio: Xudong Chen received the B.S. and M.S. degrees in electrical engineering from Zhejiang University, China, in 1999 and 2001, respectively, and the Ph.D. degree from the Massachusetts Institute of Technology, Cambridge, MA, USA, in 2005. Since 2005, he has been with the National University of Singapore. He has published 150 journal papers, with total citation 4,600+ according to ISI Web of Science (SCI). He has authored the book Computational Methods for Electromagnetic Inverse Scattering (Wiley-IEEE, 2018). His research interests mainly include electromagnetic inverse scattering, sensing and data fusion, optical/infrared/microwave scanning microscopy, optical encryption, and metamaterials. Dr. Chen was a recipient of the Young Scientist Award by the Union Radio Scientifique Internationale in 2010 and of the Best Paper Award in IEEE ICCEM conference in 2019. He has organized 20+ sessions on the topic of inverse scattering and imaging in various conferences. He has been members of organizing committees of 10+ conferences, serving as General Chair, TPC Chair, Award Committee Chair, etc. He has been an Associate Editor of the IEEE Transactions on Microwave Theory and Techniques since 2015.

Frank Dellaert

College of Computing, Georgia Institute of Technology

Title: Factor Graphs for Flexible Inference in Robotics and Vision

Abstract: In robotics and computer vision, simultaneous localization and mapping (SLAM) and structure from motion (SFM) are important and closely related problems. I will review how SLAM, SFM, and other problems in robotics and vision can be posed in terms of factor graphs, which provide a graphical language in which to develop and collaborate on such problems. The theme of the talk will be to emphasize the advantages and intuition that come with analyzing factor graphs. I will show how using these insights we have developed both batch and incremental algorithms defined on graphs in the SLAM/SFM domain, as well as more sophisticated approaches to trajectory optimization. Many of these ideas are embodied in the Skydio R1, a commercially available, fully autonomous drone I helped develop at Skydio, a San Francisco Bay area startup.

Bio: Frank Dellaert is a professor in the School of Interactive Computing at the Georgia Institute of Technology. While on leave from Tech in 2016-2018, he served as a technical project lead at Facebook Reality Labs. Before that, he completed a stint as chief scientist at Skydio, a startup founded by MIT grads to create intuitive interfaces for micro-aerial vehicles. Dellaert’s research interests lie in the overlap of robotics and computer vision, and he is particularly interested in graphical model techniques to solve large-scale problems in mapping and 3D reconstruction. The GTSAM toolbox embodies many of the ideas his research group has worked on in the past few years and is available for download at https://gtsam.org

Atilla Eryilmaz

Department of Electrical and Computer Engineering, Ohio State University

Title: Leveraging Side-Information for Learning and Optimization under Uncertainty with Applications in Social and Communication Networks

Abstract: A recurring problem in the efficient control of networked systems is the need to operate under uncertainty of critical system dynamics. It is usually necessary to develop mechanisms that optimize the target performance measures while also allocating part of their available resources to learning the uncertainties. Optimization of this, learning (exploration) - earning (exploitation) tradeoff has formed the core of many interesting approaches over the last decade that build over the multi-armed bandit (MAB) framework. An important challenge in this research space is concerned with the modeling and utilization of side-information that is typically available about other arms when an arm is pulled. Numerous applications, ranging from social to communication networks, possess different forms of side-information structures, which call for new learning and optimization mechanisms that utilize them for provably low-regret performance guarantees. In this talk, I will provide our research findings from multiple domains, including advertising in social networks, delay-constrained multi-channel communication, and multi-armed bandits for renewal processes, that reveal the gains and means of utilizing various forms of side-information in the optimal learning and control of networks under uncertainty.

Bio: Atilla Eryilmaz is a Professor of Electrical and Computer Engineering at The Ohio State University, where he has been a faculty since 2007. He received his M.S. and Ph.D. degrees in Electrical and Computer Engineering from the University of Illinois at Urbana-Champaign in 2001 and 2005, respectively. Between 2005 and 2007, he worked as a Postdoctoral Associate at the Laboratory for Information and Decision Systems (LIDS) at the Massachusetts Institute of Technology. Dr. Eryilmaz's research interests span optimal control of stochastic networks, learning, optimization theory, and information theory. He received the NSF-CAREER Award in 2010 and two Lumley Research Awards for Research Excellence in 2010 and 2015. He is a co-author of the 2012 IEEE WiOpt Conference Best Student Paper, and ore recently received the 2016 IEEE Infocom, 2017 IEEE WiOpt Conference, 2018 IEEE WiOpt, and 2019 IEEE Infocom Conference Best Paper Awards.

James Gee

Department of Radiology, Perelman School of Medicine, University of Pennsylvania

Title: Radiological Differential Diagnoses by Machine

Abstract: In this presentation of some of our work in the area of automated medical imaging diagnosis, I hope to persuade you of the value of domain expertise especially in the realm of the clinic. A quick analogy: you can conceivably build self-driving cars ignoring the fact that there exist traffic laws, and instead learn these laws from data - or you can directly encode these laws as part of your self-driving system. This is the same simple principle we champion here, and for concreteness will demonstrate within the clinical realm of neuroradiology. There have been many advances in neuroscience methods used to quantify brain MRI and similar progress in the development of various forms of artificial intelligence (AI) technology. Yet, image interpretation continues to rely nearly entirely on visual review, which has not fundamentally changed in radiology since the very first radiology report. We all recognize the opportunities to improve this situation and indeed the critical need to do so given that as more imaging studies are ordered, radiology workload continues to increase, leading to both errors and burnout. This talk will describe one instantiation of a robot radiologist that is interesting not just because it offers a real-world peek into the promise of radiology AI that has so excited our community. But, equally important, because it can be readily built from ‘off-the-shelf’ components widely available to the community.

Bio: James Gee, Ph.D., is Director of the Penn Image Computing and Science Laboratory, HHMI-NIBIB Interfaces Program in Biomedical Imaging and Informational Sciences, and the UESTC-UPenn Center for Digital Health Innovation, and Co-Director of the Translational Biomedical Imaging Center, all at the University of Pennsylvania, Philadelphia. Dr. Gee's interests are broadly in the field of biological and medical image analysis and computing. He is internationally recognized for a long track record of methodological innovation in nearly every area of the field and his commitment to translating research accomplishments into acclaimed open-source software – his group's ANTs, DTI-TK, N4ITK, ITK and ITK-SNAP software are consistently ranked as the best performing and most widely used applications in segmentation, registration, DTI analysis and morphometry. Dr Gee and his group's large portfolio of interdisciplinary collaborations spans different model systems and the major modalities in biological and medical imaging, in integrative studies of structure-function relationships of the brain, breast, eye, heart, lung, placenta and musculoskeletal system in health and disease.

Richard Hartley

College of Engineering & Computer Science, Australian National University

Title: Mathematics and Geometry, Their Role in Computer Vision in the Deep Learning Age

Abstract:

Bio: Professor Richard Hartley worked at the General Electric Research and Development Center from 1985 to 2001. In 1991 he was awarded GE's Dushman Award for his important contributions. He also worked on medical imagery with GE Medical Systems, and fingerprint imaging and aircraft engine inspection with Lockheed-Martin. In 1991, he began an extended research effort in the area of applying projective geometry techniques to reconstruction using calibrated and semi-calibrated cameras. This research direction was one of the dominant themes in computer vision research throughout the 1990s. In 2000, he co-authored (with Andrew Zisserman) a book for Cambridge University Press, summarizing the previous decade's research in this area and currently counting over 25000 citations. He has more than 50000 citations, an h-index of 74, and an i10-index of 212 on Google scholar. Professor Richard Hartley is the winner of the “Significant Computer Vision Researcher Award” at the International Conference on Computer Vision 2011, the most prestigious international conference in the domain of computer vision. He was one of the very first computer vision researchers to receive the award. He is a Fellow of the Australian Academy of Science, and a Fellow of the IEEE.

Thomas Hou

Department of Computing & Mathematical Sciences, California Institute of Technology

Title: Solving Multiscale Problems by Integrating Physical Models with Deep Learning

Abstract: In many practical applications, we often need to provide solutions to quantities of interest to a large-scale problem but with only subsampled data and partial information of the physical model. Traditional numerical methods cannot be used directly for this purpose. On the other hand, many powerful techniques have been developed in data science to represent and compress data for useful information with extreme efficiency and low computational complexities. A crucial factor for the success of these methods is to exploit some low rank or sparsity structures in these high-dimensional data. In this talk, we describe some of our recent efforts in integrating physical models and mathematical analysis with deep learning to develop a new class of numerical methods that can solve large-scale physical or data science problems by using only sub-sampled data and partial knowledge of the physical model.

Bio: Thomas Yizhao Hou is the Charles Lee Powell professor of applied and computational mathematics at Caltech. His research interests include 3D Euler singularity, interfacial flows, multiscale problems, and adaptive data analysis. He received his Ph.D. from UCLA in 1987, and joined the Courant Institute as a postdoc in 1987. He became a tenure track assistant professor at the Courant Institute in 1989 and then was promoted to tenured associate professor in 1992. He moved to Caltech in 1993, served as the department chair of applied and computational mathematics from 2000 to 2006, and was named the Charles Lee Powell Professor in 2004. Dr. Hou is a Fellow of American Academy of Arts and Sciences, Society of Industrial and Applied Mathematics (SIAM), and American Mathematical Society. He was also the founding Editor-in-Chief of the SIAM Journal on Multiscale Modeling and Simulation from 2002 to 2007.

Jianwei Huang

School of Science and Engineering, The Chinese University of Hong Kong, Shen Zhen

Title: Economics of Mobile Crowd Sensing

Abstract: Mobile crowd sensing achieves a flexible and scalable sensing coverage with a low deploying cost, by encouraging mobile users to participate and contribute their smartphones as sensors. In this talk, we first provide an overview regarding the core research issues in mobile crowd sensing. Then we focus on the design of effective reward systems to induce high quality contributions by users, considering the diversity, social relationship, and bounded rational decision process of users.

Bio: Jianwei Huang is a Presidential Chair Professor and the Associate Dean of the School of Science and Engineering, The Chinese University of Hong Kong, Shenzhen. He is also a Professor in the Department of Information Engineering, The Chinese University of Hong Kong. He received the Ph.D. degree from Northwestern University in 2005, and worked as a Postdoc Research Associate at Princeton University during 2005-2007. He has been an IEEE Fellow, a Distinguished Lecturer of IEEE Communications Society, and a Clarivate Analytics Highly Cited Researcher in Computer Science. He is the co-author of 9 Best Paper Awards, including IEEE Marconi Prize Paper Award in Wireless Communications in 2011. He has co-authored six books, including the textbook on "Wireless Network Pricing." He received the CUHK Young Researcher Award in 2014 and IEEE ComSoc Asia-Pacific Outstanding Young Researcher Award in 2009. He has served as an Associate Editor of IEEE Transactions on Mobile Computing, IEEE/ACM Transactions on Networking, IEEE Transactions on Network Science and Engineering, IEEE Transactions on Wireless Communications, IEEE Journal on Selected Areas in Communications - Cognitive Radio Series, and IEEE Transactions on Cognitive Communications and Networking. He has served as the Chair of IEEE ComSoc Cognitive Network Technical Committee and Multimedia Communications Technical Committee. He is the recipient of IEEE ComSoc Multimedia Communications Technical Committee Distinguished Service Award in 2015 and IEEE GLOBECOM Outstanding Service Award in 2010. More detailed information can be found at http://jianwei.ie.cuhk.edu.hk/.

Longbo Huang

Institute for Interdisciplinary Information Sciences, Tsinghua University

Title: Multi-armed Bandits with Compensation

Abstract: We propose and study the known-compensation multi-arm bandit (KCMAB) problem, where a system controller offers a set of arms to many short-term players for T steps. In each step, one short-term player arrives at the system. Upon arrival, the player aims to select an arm with the current best average reward and receives a stochastic reward associated with the arm. In order to incentivize players to explore other arms, the controller provides a proper payment compensation to players. The objective of the controller is to maximize the total reward collected by players while minimizing the compensation. We first provide a compensation lower bound Θ(∑_i(Δ_i log⁡T)/(KL)_i ), where Δ_i and (KL)_i are the expected reward gap and Kullback-Leibler (KL) divergence between distributions of arm i and the best arm, respectively. We then analyze three algorithms to solve the KCMAB problem, and obtain their regrets and compensations. We show that the algorithms all achieve O(log⁡T) regret and O(log⁡T) compensation that match the theoretical lower bound. Finally, we present experimental results to demonstrate the performance of the algorithms.

Bio: Dr. Longbo Huang is an associate professor (with tenure) at the Institute for Interdisciplinary Information Sciences (IIIS) at Tsinghua University, Beijing, China. He received his Ph.D. in EE from the University of Southern California, and then worked as a postdoctoral researcher in the EECS dept. at University of California at Berkeley before joining IIIS. Dr. Huang currently serves as an editor for IEEE Transactions on Communications (TCOM), an associate editor for ACM Transactions on Modeling and Performance Evaluation of Computing Systems (ToMPECS), and an associate editor for IEEE/ACM Transactions on Networking (ToN). He is a senior member of IEEE and a member of ACM. Dr. Huang has held visiting positions at the LIDS lab at MIT, the Chinese University of Hong Kong, Bell-labs France, and Microsoft Research Asia (MSRA). He was a visiting scientist at the Simons Institute for the Theory of Computing at UC Berkeley in Fall 2016. Dr. Huang received the Outstanding Teaching Award from Tsinghua university in 2014. He received the Google Research Award (co-recipient) and the Microsoft Research Asia Collaborative Research Award in 2014, and was selected into the MSRA StarTrack Program in 2015. Dr. Huang won the ACM SIGMETRICS Rising Star Research Award in 2018. Dr. Huang’s current research interests are in the areas of stochastic modeling and analysis, reinforcement learning and control, optimization and machine learning, and big data analytics.

Sheng-Kwang Hwang

Department of Photonics, National Cheng Kung University

Title: Photonic Microwave Processing Using Nonlinear Dynamics of Semiconductor Lasers for Radio-over-fiber Applications

Abstract: Radio-over-fiber (RoF) has attracted great attention due to the strong demand in distributing microwave subcarriers over long distances through fibers for antenna remoting applications, such as broadband wireless access networks in the next generation wireless communication. External or direct modulation of semiconductor lasers is the simplest scheme to superimpose microwave subcarriers onto optical carriers for such RoF systems. However, for many RoF applications, processing of such generated microwave subcarriers is necessary in order to, for example, counteract microwave power fading while distributing the microwave subcarriers over fibers, improve detection sensitivity and transmission distance after photodetection, recover the microwave subcarriers for coherent detection, or introduce time delays to the microwave subcarriers for phased array antennas. In this talk, I will demonstrate how a semiconductor laser operating at nonlinear dynamics can be used as an all-optical and multi-purpose microwave processing device for RoF applications.

Bio: Sheng-Kwang Hwang received the B.S. degree in electro-physics from National Chiao Tung University, Hsinchu, Taiwan, in 1993, and M.S. and Ph.D. degrees in electrical engineering from the University of California, Los Angeles, in 1999 and 2003, respectively. He was an assistant professor with the Graduate Institute of Opto-Mechatronics, National Chung Cheng University, Taiwan, from 2003 to 2007. He was an assistant professor and an associate professor with the Department of Photonics, National Cheng Kung University, Taiwan, from 2007 to 2009 and from 2009 to 2015, respectively, and he has been a professor with the same department since 2015. His current research interests include semiconductor lasers, nonlinear dynamics, optical communications, optical signal processing, microwave photonics, and radio-over-fiber. He received the Dr. Bor-Uei Chen Scholarship of the Photonics Society of Chinese-Americans, USA in 2001, and the Y. Z. Hsu Science Paper Award of the Far Eastern Y. Z. Hsu Science and Technology Memorial Foundation, Taiwan in 2018. He served on the technical program committee of IEEE International Topical Meetings on Microwave Photonics (2015, 2016, 2018), Progress in Electromagnetics Research Symposium (2018), Microoptics Conference (2018), Wireless and Optical Communications Conference (2018), and Optics and Photonics Taiwan, International Conference (2011~2018). He also served on the organizing committee of International Symposium on Physics and Applications of Laser Dynamics (2013, 2014, 2015, 2017, 2018), and as an organizing co-chair of International Symposium on Physics and Applications of Laser Dynamics (2011, 2012, 2016).

Nan Jiang

Department of Computer Science, University of Illinois at Urbana-Champaign

Title: Sample-efficient Exploration in Reinforcement Learning with Function Approximation

Abstract: In reinforcement learning (RL), autonomous agents solve sequential decision-making problems by (1) actively exploring the environment to collect data, and (2) improving behavior by learning from the collected data. The recent success of RL can largely be attributed to the use of advanced function approximation techniques (e.g., deep neural networks) for the learning component. In contrast, advances in exploration techniques have been rather limited: most existing algorithms that perform systematic exploration only apply to simple problems with small state spaces, and cannot accommodate sophisticated function approximation schemes that are critical for solving real-world problems with rich observations.

Bio: Nan Jiang is an assistant professor in Department of Computer Science at University of Illinois at Urbana-Champaign. Prior to joining UIUC, he was a postdoc researcher at Microsoft Research, New York City. He received his PhD in Computer Science and Engineering at University of Michigan, advised by Satinder Singh. His research interests lie in theory of reinforcement learning, mostly focusing on sample efficiency. Specific research topics include exploration theory for function approximation, state representation learning, off-policy evaluation, spectral learning of dynamical systems, etc. He is recipient of the Best Paper Award at AAMAS 2015 and Rackham Predoctoral Fellowship in 2016.

Tao Jiang

Department of Computer Science and Engineering, University of California

Title: Predicting Isoform Functions from Sequences and Expression Profiles via Deep Learning

Abstract: Alternative splicing generates multiple isoforms from a single gene, greatly increasing the functional diversity of a genome. Although gene functions have been well studied, little is known about the specific functions of isoforms, making accurate prediction of isoform functions highly desirable. However, the existing approaches to predicting isoform functions are far from satisfactory due to at least two reasons: i) Unlike genes, isoform-level functional annotations are scarce and ii) the information of isoform functions is concealed in various types of data including isoform sequences, co-expression relationship among isoforms, etc. In this study, we present a novel approach, DIFFUSE, to predict isoform functions. To integrate various types of data, our approach adopts a hybrid framework by first using a deep neural network (DNN) to predict the functions of isoforms from their genomic sequences and then refining the prediction using a conditional random field (CRF) based on co-expression relationship. To overcome the lack of isoform-level ground truth labels, we further propose an iterative semi-supervised learning algorithm to train both the DNN and CRF together. Our extensive computational experiments demonstrate that DIFFUSE could effectively predict the functions of isoforms and genes with an accuracy significantly higher than the state-of-the-art methods. We further validate the prediction results by analyzing the correlation between functional similarity, sequence similarity, expression similarity, and structural similarity, as well as the consistency between the predicted functions and well-studied functional features of isoform sequences. This is a joint work with Hao Chen, Dipan Shaw, Jianyang Zeng and Dongbo Bu.

Bio: Tao Jiang received B.S. in Computer Science and Technology from the University of Science and Technology of China, Hefei, in July 1984 and Ph.D. in Computer Science from University of Minnesota in Nov. 1988. He was a faculty member at McMaster University, Hamilton, Ontario, Canada during Jan. 1989 - July 2001 and is now Professor of Computer Science and Engineering at University of California - Riverside (UCR). He is also a member of the UCR Institute for Integrative Genome Biology, a member of the Center for Plant Cell Biology, a principal scientist at Shanghai Center for Bioinformation Technology, and Qianren Chair Visiting Professor at Tsinghua University. Tao Jiang's recent research interest includes combinatorial algorithms, computational molecular biology, bioinformatics, and computational aspects of information retrieval. He is a fellow of the Association for Computing Machinery (ACM) and of the American Association for the Advancement of Science (AAAS), and held a Presidential Chair Professor position at UCR during 2007-2010. He has published over 280 papers in computer science and bioinformatics journals and conferences, and won several best paper awards. More information about his work can be found at http://www1.cs.ucr.edu/~jiang

Xiangyan Kong

Shanghai Institute of Microsystem and Information Technology, CAS

Title: Improvement of 36-channel Magnetocardiography System Based on DC SQUID Magemetometers

Abstract: The SQUID-based biomagnetic measurement systems are widely used in the diagnosis of heart and brain diseases [1]. In SIMIT, we set up the first multi-channel MCG system based on the weakly damped DC SQUID magnetometers with average field noise of 5 fT/Hz [2, 3]. The system consists of 36 sensing magnetometers which mainly measure the cardiac magnetic field and several reference magnetometers which measure the environmental magnetic field, thus realizing software gradiometer configurations. The sensor coverage area of the 36 channels is 2020 cm2 and the baseline of software gradiometer is 6 cm. The bandwidth of the magnetometers is larger than 3.6 kHz and the detective sensitivity of the magnetometers is lower than 0.05 pT. In addition, a new design of dewar insert with liquid nitrogen interlayer was adopted to prolong the duration of liquid helium. The average boil-off rate of liquid helium is 0.22 L/h with the system continuously running. MCG signal with SNR of 40 dB in different magnetic shielded rooms could be obtained from our system. [1] K. Kazami, Y. Takada, S. Fujimoto, T. Yoshida, H. Ogata, and H. Kado, “A Drung-type magnetometer mounted on a GM cryocooler,” Superconductor Science and Technology, vol. 7, no. 5, pp. 256-259, May, 1994. [2]C. Liu, Y. Zhang, M. Mück, S. Zhang, H. Krause, A. Braginski, G. Zhang, Y. Wang, X. Kong, X. Xie, A. Offenhäusser and M. Jiang. “Statistical characterization of voltage-biased SQUIDs with weakly-damped junctions,” Superconductor Science andTechnology, vol. 26, no. 06, 065002, 2013. [3] Y. Qiu, H. Li, S. Zhang, Y. Wang, X. Kong, C. Zhang, Y. Zhang, X. Xu, K. Yang and X. Xie, “Low-Tc direct current superconducting quantum interference devices magnetometer-based 36-channel magnetocardiography system in a magnetically shielded room,” Chinese Physics B, vol. 24, no. 7, 078501, 2015.

Bio:

Puxiang Lai

Department of Biomedical Engineering, The Hong Kong Polytechnic University

Title: Wavefront Shaping-enabled Optical Focusing and Its Application Towards Deep-tissue Single Neuron Stimulation

Abstract: Light, in many ways, is an ideal form of electromagnetic waves to probe and treat biological tissues. Its applications, however, encounter an inevitable trade-off between resolution and penetration depth due to the strong scattering of light in tissue. Existing microscopic optical modalities seldom can see beyond the so-called optical diffusion limit (~1 mm for human skin). In this talk, we present our endeavors in the past years to achieve high-resolution optical focusing within or through thick biological tissue or tissue-like scattering media via optical wavefront shaping. We also show our preliminary results of applying wavefront shaping implementations towards temporal evolutional optogenetics with single neuron precision through brain skull. Further direction is also discussed.

Bio: Dr. Puxiang Lai received his Bachelor of Engineering in Biomedical Engineering from Tsinghua University, China in 2002, Master of Science in Acoustics from the Chinese Academy of Sciences, China in 2005, and PhD in Mechanical Engineering from the Boston University, USA in 2011. After that, Dr. Lai joined Dr Lihong Wang’s Optical Imaging Lab in the Washington University in St. Louis as a Postdoctoral Research Associate. In September 2015, he joined the Department of Biomedical Engineering at the Hong Kong Polytechnic University as tenure-tracked Assistant Professor. Dr. Lai’s research focuses on the development of novel biomedical imaging, therapy, and manipulation modalities by using light and sound. Current research interests include, but are not limited to, optical wavefront shaping, photoacoustic imaging, acousto-optic imaging, optogenetics, as well as application of artificial intelligence techniques in biomedical diagnosis and imaging. Thus far, his research has fueled more than 30 top journal publications (including two first-authored in Nature Photonics and Nature Communications, respectively), more than 60 invited or contributing talks at important international conferences with several best paper/poster awards. Dr. Lai has also been invited to give seminars in many famous institutions, and served as editor or reviewer for more than 25 premium journals. Since joining BME at PolyU in late 2015, Dr. Lai aims to continue his research on deep-tissue probing and treatment by using light and sound, which has been supported by the Hong Kong Polytechnic University, the Hong Kong Research Grants Council (RGC), the Hong Kong Innovation and Technology Commission (ITC), the National Natural Science Foundation of China (NSFC), and the Shenzhen Science and Technology Innovation Commission (STIC). He has been awarded the 2016-2017 Hong Kong RGC Early Career Award, and the Hong Kong Polytechnic University Faculty of Engineering Research Grant Achievement Award.

Lawrence Le

Department of Radiology and Diagnostic Imaging, University of Alberta

Title: Ultrasonic Imaging of Cortical Bones

Abstract: Osteoporosis is a systemic disorder of the skeleton affecting over 200 million women worldwide. The bone disease is characterized by low bone mass, deteriorated microstructures, and cortical thinning of bone tissues. The loss of trabeculae and cortical bones through the process of trabecularization disrupts the bone strength and consequently leads to increase of bone fragility and fracture risk. The rate of bone loss is greater in women than men with women experiencing accelerated bone loss following menopause. Cortical thickness and elasticity are important determinants of bone strength. The cortex is a strong waveguide because it has much larger acoustic impedance than those of the surrounding soft tissues. Quantitative guided wave ultrasonography is particularly attractive because of the sensitivity of guided waves to the geometry, micro-architecture, and material properties of the cortex. However, the problem to extract bone properties of the cortex using guided waves is challenging because the long bone is inhomogeneous, absorptive, and transversely isotropic with irregular thickness and surface. In this presentation, I will present some of the methodology that our group has developed to solve the problem.

Bio: Lawrence H. Le received his PhD in earth physics from the University of Alberta, Edmonton, Canada. He held a NSERC (Natural Sciences and Engineering Research Council of Canada) industrial postdoctoral fellowship in Schlumberger-Doll Research Lab, Ridgefield, Connecticut. He started his medical physics residency training in the Department of Radiology and Diagnostic Imaging (DRDI), University of Alberta in 1994. Subsequently, he completed a MBA degree in Finance and Technology commercialization at the University of Alberta. He joined the DRDI at the University of Alberta as an academic staff and Capital Health as a clinical medical physicist in 2000. He is currently a clinical professor leading the graduate program in DRDI and the Edmonton Authorized Radiation Protection Agency within Alberta Health Services. He is also a senior Visiting Scholar of the State Key Laboratory of ASIC and System, Fudan University. He directs the Ultrasonic Bone Tissue Characterization and Imaging group. He guides his students to use vigorous geophysical techniques to image and study bone tissues. His research interests are in imaging, signal and image processing, simulation, and inversion. Lawrence is a member of AAPM (The American Association of Physicists in Medicine) and COMP (The Canadian Organization of Medical Physicists).

Xiuling Li

Department of Electrical and Computer Engineering, University of Illinois at Urbana-Champaign

Title: Extreme Miniaturization of Passive Electronic Components by Self-Rolled-up Membrane Nanotechnology

Abstract: Inductors and transformers remain important fundamental components of many RF circuits, including low and band pass filters, wireless power transfer and telemetry, and other system-in-package (SiP) architectures. These applications demand smaller footprint and high efficiency. The overarching physical principle of self-rolled-up membrane (S-RuM) nanotechnology is strain-driven spontaneous deformation of 2D membranes into 3D complex architectures. S-RuM inductors have been demonstrated with a footprint that is 10 – 100 times smaller than their 2D counterparts and capable of operating at high frequencies (> 20 GHz). By stacking two S-RuM inductors in-plane or vertically to form transformers, near unity coupling coefficients and unprecedentedly high turns ratios have been achieved. Through global and local strain engineering, S-RuM filters, transmission lines, antennas with ultra-high frequency and bandwidth can all be enabled. S-RuM nanotechnology promises to break the constraints of size, weight, and performance (SWAP) of RFICs and even millimeter wave communications. The unique form factor can also bring translational impact to wearable and flexible IoT devices and other exciting opportunities.

Bio: Xiuling Li received her B.S. degree form Peking University and Ph.D. degree from the University of California at Los Angeles. Following post-doctoral positions at California Institute of Technology and University of Illinois, as well as industry experience at EpiWorks, Inc., she joined the faculty of the University of Illinois in 2007 in the Department of Electrical and Computer Engineering. She was promoted to Associate Professor with tenure in 2012, and to Professor and Willett Faculty Scholar in 2015. She has published >140 journal papers and holds >20+ patents in the area of semiconductor materials and devices, with a Google Scholar h-index of 46. She is a Fellow of IEEE and APS. Her other honors and awards include NSF CAREER award, DARPA Young Faculty Award, ONR Young Investigator Award. She currently serves as a Deputy Editor of Applied Physics Letters and Vice President for Finance and Administration of the IEEE Photonics Society.

Yixue Li

Department of Bioinformatics & Biostatistics, Shanghai Jiaotong University

Title: Genomics study of animal models

Abstract: Animal models provide us with a great tool to study important and basic biological functions. As pointed out by August Krogh：“For such a large number of problems there will be some animal of choice, or a few such animals, on which it can be most conveniently studied”。 Through the genome research of camels, dogs and rabbits, we have discovered some important metabolic pathways, functional genes and corresponding genetic variation sites, which can explain the causes and protection mechanisms of some important biological processes. It provides a way for us to further study the mechanism of disease occurrence and development.

Bio: Yixue Li, Research fellow, Ph.D. supervisor , director of the Shanghai Society for Bioinformatics, head of Specialist Group in the bioinformatics technology of biological and agricultural technology field for 863 Program during the 10th Five-Year Plan, specialist in the biological and pharmaceutical technology field for 863 Program during the 11th Five-Year Plan. Dr. Li received his Ph.D. degree in theoretical physics from the University of Heidelberg, Germany, in October 1996. After Dr. Li got his Ph.D. degree he worked as a bioinformatics research staff in European Molecular Biology Laboratory (EMBL) from April 1997 to June 2000, and he came back to Shanghai, China in July 2000. In July 2000, he served as the deputy director of the bioinformation center, SIBS , CAS and in 2002 served as the director. Since July 2002 he hold a concurrent post as the director of Shanghai Center for Bioinformation Technology, and since 2007 he hold a concurrent post as the director of the Department of Bioinformatics and Biostatistics of Shanghai Jiaotong University. In January 2007, he served as the vice director of Key Laboratory of Systems Biology, SIBS, CAS. In January 2015, he served as the research group leader of the Institute of Biochemistry and Cell Biology, SIBS, CAS. Currently he is the director of the Bio-Med Big Data Center of PICB, SIBS , CAS.

Ming Lin

Department of Computer Science, University of Maryland

Title: Reconstructing Reality: From Physical World to Virtual Environments

Abstract: With increasing availability of data in various forms from images, audio, video, 3D models, motion capture, simulation results, to satellite imagery, representative samples of the various phenomena constituting the world around us bring new opportunities and research challenges. Such availability of data has led to recent advances in "data-driven modeling” that transforms the data captured in the physical world to a digital replica of the real world in a virtual environment. In this talk, I survey recent advances that integrate classical model-based methods and statistical learning techniques to tackle challenging problems that have not been previously addressed. These approaches offer new insights for understanding complex collective behaviors, developing better models for complex dynamical systems from captured data, delivering more effective medical diagnosis and treatment, as well as cyber-manufacturing of customized apparel. I conclude by discussing some possible future directions and challenges.

Bio: Ming C. Lin is currently the Elizabeth Stevinson Iribe Chair of Computer Science at the University of Maryland College Park and John R. & Louise S. Parker Distinguished Professor Emerita of Computer Science at the University of North Carolina (UNC), Chapel Hill. She was also an Honorary Visiting Chair Professor at Tsinghua University in China and at University of Technology Sydney in Australia. She obtained her B.S., M.S., and Ph.D. in Electrical Engineering and Computer Science from the University of California, Berkeley. She received several honors and awards, including the NSF Young Faculty Career Award in 1995, Honda Research Initiation Award in 1997, UNC/IBM Junior Faculty Development Award in 1999, UNC Hettleman Award for Scholarly Achievements in 2003, Beverly W. Long Distinguished Professorship 2007-2010, UNC WOWS Scholar 2009-2011, IEEE VGTC Virtual Reality Technical Achievement Award in 2010, and many best paper awards at international conferences. She is a Fellow of ACM, IEEE, and Eurographics. Her research interests include computational robotics, haptics, physically-based modeling, virtual reality, sound rendering, and geometric computing. She has (co-)authored more than 300 refereed publications in these areas and co-edited/authored four books. She has served on hundreds of program committees of leading conferences and co-chaired dozens of international conferences and workshops. She is currently a member of Computing Research Association-Women (CRA-W) Board of Directors, Chair of IEEE Computer Society (CS) Transactions Operating Committee, Chair of IEEE CS Computer Pioneer Award, and Chair of ACM SIGGRAPH Outstanding Doctoral Dissertation Award. She is a former member of IEEE CS Board of Governors, a former Editor-in-Chief of IEEE Transactions on Visualization and Computer Graphics (2011-2014), a former Chair of IEEE CS Transactions Operations Committee, and a member of several editorial boards. She also has served on several steering committees and advisory boards of international conferences, as well as government and industrial technical advisory committees.

Jia Liu

Department of Computer Science, Iowa State University

Title: Combinatorial Bandits with Fairness Constraints

Abstract: The multi-armed bandit (MAB) model has been widely adopted for studying many practical optimization problems (network resource allocation, ad placement, crowdsourcing, etc.) with unknown parameters. The goal of the player (i.e., the decision maker) here is to maximize the cumulative reward in the face of uncertainty. However, the basic MAB model neglects several important factors of the system in many real-world applications, where multiple arms (i.e., actions) can be simultaneously played and an arm could sometimes be “sleeping” (i.e., unavailable). Besides reward maximization, ensuring fairness is also a key design concern in practice. To that end, we propose a new Combinatorial Sleeping MAB model with Fairness constraints, called CSMAB-F, aiming to address the aforementioned crucial modeling issues. The objective is now to maximize the reward while satisfying the fairness requirement of a minimum selection fraction for each individual arm. To tackle this new problem, we extend an online Upper Confidence Bound (UCB) algorithm to deal with a critical tradeoff between exploitation and exploration and employ the virtual queue technique to properly handle the fairness constraints. By carefully integrating these two techniques, we develop a new algorithm, called Learning with Fairness Guarantee (LFG), for the CSMAB-F problem. Further, we rigorously characterize LFG’s feasibility-optimality and regret upper bound. Finally, we perform extensive simulations to corroborate the effectiveness of the proposed algorithm. Interestingly, the simulation results reveal an important tradeoff between the regret and the speed of convergence to a point satisfying the fairness constraints.

Bio: Jia (Kevin) Liu is currently an Assistant Professor in the Dept. of Computer Science and Dept. of Electrical and Computer Engineering (by courtesy) at Iowa State University, where he joined in Aug. 2017. He received his Ph.D. degree from the Bradley Dept. of Electrical and Computer Engineering at Virginia Tech in 2010. He was a Postdoctoral Researcher and subsequently a Research Assistant Professor from Feb. 2010 to Jul. 2017, both in the Dept. of Electrical and Computer Engineering at The Ohio State University. His research areas include theoretical foundations of control and optimization for stochastic networked systems, distributed algorithms design, optimization of cyber-physical systems, data analytics infrastructure, and machine learning. Dr. Liu is a senior member of IEEE and a member of ACM. His work has received numerous awards at top venues, including IEEE INFOCOM'19 Best Paper Award, IEEE INFOCOM'16 Best Paper Award, IEEE INFOCOM'13 Best Paper Runner-up Award, IEEE INFOCOM'11 Best Paper Runner-up Award, and IEEE ICC'08 Best Paper Award. He is a recipient of Bell Labs President Gold Award in 2001. His research has been supported by NSF, AFOSR, AFRL, and ONR.

Jiaying Liu

Institute of Computer Science & Technology, Peking University

Title: Intelligent Visual Computing

Abstract: Intelligent image/video computing is a fundamental topic in image processing which has witnessed rapid progress in the last two decades. Due to various degradations in the image and video capturing, transmission and storage, image and video include many undesirable effects, such as low resolution, low light condition, rain streak and rain drop occlusions. The recovery of these degradations is ill-posed. With the wealth of statistic-based methods and learning-based methods, this problem can be unified into the cross-domain transfer, which cover more tasks. I will also share some recent progresses of low level vision tasks in my group.

Bio: Jiaying Liu is currently an Associate Professor with the Institute of Computer Science and Technology, Peking University. She received the Ph.D. degree (Hons.) in computer science from Peking University, Beijing China, 2010. She has authored over 100 technical articles in refereed journals and proceedings, and holds 31 granted patents. Her current research interests include multimedia signal processing, compression, and computer vision. Dr. Liu is a Senior Member of IEEE and CCF. She was a Visiting Scholar with the University of Southern California, Los Angeles, from 2007 to 2008. She was a Visiting Researcher with the Microsoft Research Asia in 2015 supported by the Star Track Young Faculties Award. She has served as a member of the Multimedia Systems & Applications Technical Committee (MSA-TC), Visual Signal Processing and Communications Technical Committee (VSPC-TC) and the Education and Outreach Technical Committee (EO-TC) in IEEE Circuits and Systems Society, a member of the Image, Video, and Multimedia (IVM) Technical Committee in APSIPA. She has also served as the Technical Program Chair of IEEE VCIP-2019/ACM ICMR-2021, the Publicity Chair of IEEE ICIP-2019/VCIP-2018, the Grand Challenge Chair of IEEE ICME-2019, and the Area Chair of ICCV-2019. She was the APSIPA Distinguished Lecturer (2016-2017).

Ling Liu

School of Computer Science, Georgia Institute of Technology

Title: Robustness of Deep Learning Systems against Adversarial Inputs

Abstract: We are entering an exciting era where human intelligence is being enhanced by machine intelligence through artificial intelligence (AI), machine learning (ML), big data and computer vision. However, prediction models trained using in-house training dataset suffer from two well-known problems: (1) The prediction accuracy of a trained model heavily depends on the generalization of the training dataset and may suffer from poor accuracy for unseen data inputs. (2) A privately trained prediction model is vulnerable to adversarial inputs, which manipulates the prediction outputs with only a black box access to a learning as a service API, turning a learning system against itself through input data poisoning attacks. In this talk, I will describe the formal metrics to quantitatively evaluate and measure the robustness of a trained prediction model against unseen inputs in the presence of different adversarial settings and share some of our current research results. An important takeaway message is that the defense mechanisms for guarding the robustness of a deep learning system should be geared towards improving the generalization properties of the target learning system.

Bio: Prof. Dr. Ling Liu is a Professor in the School of Computer Science at Georgia Institute of Technology. She directs the research programs in Distributed Data Intensive Systems Lab (DiSL), examining various aspects of large-scale data intensive systems. Prof. Liu is an internationally recognized expert in the areas of Big Data Systems and Analytics, Distributed Systems, Database and Storage Systems, Internet Computing, Privacy, Security and Trust. Prof. Liu has published over 300 international journal and conference articles, and is a recipient of the best paper award from a number of top venues, including ICDCS 2003, WWW 2004, Pat Goldberg Memorial Best Paper Award 2005, IEEE CLOUD 2012, IEEE ICWS 2013, ACM/IEEE CCGrid 2015, IEEE Edge 2017, IEEE ICIOT 2017. Prof. Liu has served as general chair and PC chairs of numerous IEEE and ACM conferences in the fields of big data, cloud computing, data engineering, distributed computing, very large databases, World Wide Web, and served as the editor in chief of IEEE Transactions on Services Computing from 2013-2016. Currently Prof. Liu is co-PC chair of The Web 2019 (WWW 2019) and the Editor in Chief of ACM Transactions on Internet Technology (TOIT). Prof. Liu is an elected IEEE Fellow and a recipient of IEEE Computer Society Technical Achievement Award 2012.

Si Liu

School of Computer Science and Engineering, Beihang University

Title: Human-centric Image Analysis

Abstract: In this talk, I will introduce our recent research on human-centric image analysis, including face, human and human-centric relation segmentation. (1) We develop a real-time face parsing system and a GAN based face editing method. (2) We propose a surveillance video parsing method and a domain adaptation parsing technique for real applications. (3) We define a new problem of relation segmentation. We collect a large Person in Context (PIC) dataset and manually annotate the human and object mask as well as their relations. PIC 1.0 dataset is released to the public in an ECCV 2018 challenge. We hope it will receive more attention from the academia and industry.

Bio: Si Liu is an associate professor in Beihang University. She received PHD degree from Institute of Automation, Chinese Academy of Sciences. She has been Research Assistant and Postdoc in National University of Singapore. She was a visiting scholar of Microsoft Research Asia. She has published over 40 cutting-edge papers on image editing and segmentation on TPAMI, IJCV, CVPR, ECCV and ICCV. She was the recipient of Best Paper of ACM MM 2013, Best demo award of ACM MM 2012. She was the Champion of CVPR 2017 Look Into Person Challenge. She is the organizer of ECCV 2018 and ICCV 2019 Person in ContextWorkshop/challenge. She servers as an area chair of ICCV 2019, SPC of IJCAI 2019 and AAAI 2019.

John C.S. Lui

Department of Computer Science and Engineering, The Chinese University of Hong Kong

Title: An Online Learning Approach to Network Application Optimization with Guarantee

Abstract: Network application optimization is essential for improving the performance of the application as well as its user experience. The network application parameters are crucial in making proper decisions for network application optimizations. However, many existing works are impractical by assuming a priori knowledge of the parameters which are usually unknown and need to be estimated. There have been studies that consider optimizing network application in an online learning context using multi-armed bandit models. However, existing frameworks are problematic as they only consider to find the optimal decisions to minimize the regret, but neglect the constraints (or guarantee) requirements which may be excessively violated. In this work, we first propose a novel online learning framework for network application optimizations with guarantee. To the best of our knowledge, we are the first to formulate the stochastic constrained multi-armed bandit model with time-varying “multi- level rewards” by taking both “regret” and “violation” into consideration. We are also the first to design a constrained bandit policy, Learning with Minimum Guarantee (LMG), with provable sub-linear regret and violation bounds. We illustrate how our framework can be applied to several emerging network application optimizations, namely, (1) opportunistic multichannel selection, (2) data-guaranteed crowdsensing, and (3) stability-guaranteed crowdsourced transcoding. To show the effectiveness of LMG in optimizing these applications with different minimum requirements, we also conduct extensive simulations by compar- ing LMG with existing state-of-the-art policies.

Bio: John C.S. Lui is currently the Choh-Ming Li Chair Professor in the CSE Department at The Chinese University of Hong Kong (CUHK). He received his Ph.D. in Computer Science from UCLA. His current research interests are in network sciences with large data implications, machine learning on large data analytics, network/system/mobile security, network economics, large scale distributed systems and performance evaluation theory. Currently, John is the senior editor in the IEEE/ACM Transactions on Networking, and has been serving in the editorial board of IEEE Transactions on Mobile Computing, ACM Transactions on Modeling and Performance Evaluation of Computing Systems, IEEE Transactions on Computers, IEEE Transactions on Parallel and Distributed Systems, Journal of Performance Evaluation, Journal of Network Science and International Journal of Network Security. He is a member of the review panel in the IEEE Koji Kobayashi Computers and Communications Award committee, and has served at the IEEE Fellow Review Committee. He received various departmental teaching awards and the CUHK Vice-Chancellor's Exemplary Teaching Award, as well as the CUHK Faculty of Engineering Research Excellence Award. He is an elected member of the IFIP WG 7.3, Fellow of ACM, Fellow of IEEE, Senior Research Fellow of the Croucher Foundation and was the past chair of the ACM SIGMETRICS (2011-2015). His personal interests include films and general reading.

Patrick Luk

Power Engineering Centre, Cranfield University

Title: Towards a New Energy Eco-system: The Convergence of the Electric Grid and The Electric Vehicle

Abstract: This talk will start with significant challenges that the electric grid and the automotive industry are facing today within the context of the EU Energy Vision 2050’s ‘energy trilemma’, and highlight some compelling evidence that these two historically independent sectors have now reached the right time and conditions for the convergence into a new energy ecosystem. Using results from both established and on-going research projects, the talk will discuss how enabling technologies, such as information technology and cloud computing, and disruptive applications, such as mobility as a service (MaaS), have turned some of these challenges into opportunities. It will highlight Cranfield University’s new Green Grid-Vehicle Ecosystem initiative that projects the vision of a fully de-carbonized electric grid supporting the entire transportation network and associated mobility services. The talk will finish with a futuristic electric propulsion concept based on superconducting machines for full electric aircraft beyond 2050.

Bio: Patrick Chi-Kwong Luk is Full Professor in Electrical Engineering and Head of Electric Power and Drives Group at Cranfield University, U.K. He has been the principal investigator for the successful delivery of two UK-government funded strategic projects in grid-connected electric vehicles. He is responsible for providing academic leadership and strategic direction for More Electric Technologies across the University’s different disciplines, including energy, automotive, aerospace and water. He is a member of the University’s £9M ‘Multi User Environment for Autonomous Vehicle Innovation’ government funded initiative to develop green mobile technologies built on a ‘smart’ road across the University’s campus. Currently a Distinguished Lecturer for the IEEE Vehicular Technology Society (2018-2020), Prof Luk has held technical advisory roles to blue-chip companies including Mitsubishi Electric UK, Lotus Engineering, Ministry of Defence UK, BAE Systems, and Lockheed Martin. He is the co-holder of 10 GB/US patents and applications on electric drives, and has over 250 publications and several book chapters in electric machines and power electronics.

Ke Ma

Department of Electrical Engineering, Shanghai Jiaotong University

Title: Mission Profile Based Reliability Analysis of Power Electronics System

Abstract:

Bio: Ke Ma received the B.Sc. and M.Sc. degrees in electrical engineering from the Zhejiang University, China in 2007 and 2010 respectively. He received the Ph.D. degree from the Aalborg University, Denmark in 2013, where he was a Postdoc in 2013 and became an Assistant Professor in 2014. He was part-time consultant with Vestas Wind Systems A/S, Denmark in 2015. In 2016 he joined the faculty of Shanghai Jiao Tong University, China as a tenure-track Research Professor. His current research interests include the power electronics and its reliability in the application of HVDC, renewable energy and motor drive systems. He is now an IEEE senior member, and is serving as Associate Editors for two IEEE journals. He was the receiver of “Excellent Young Wind Doctor Award 2014” by European Academy of Wind Energy, and several prized paper awards by IEEE.

Dinesh Manocha

Department of Computer Science, University of Maryland

Title: Autonomous Driving: Simulation and Navigation

Abstract: Autonomous driving has been an active area of research and development over the last decade. Despite considerable progress, there are many open challenges including automated driving in dense and urban scenes. In this talk, we give an overview of our recent work on simulation and navigation technologies for autonomous vehicles. We present a novel simulator, AutonoVi-Sim, that uses recent developments in physics-based simulation, robot motion planning, game engines, and behavior modeling. We describe novel methods for interactive simulation of multiple vehicles with unique steering or acceleration limits taking into account vehicle dynamics constraints. In addition, AutonoVi-Sim supports navigation for non-vehicle traffic participants such as cyclists and pedestrians AutonoVi-Sim also facilitates data analysis, allowing for capturing video from the vehicle's perspective, exporting sensor data such as relative positions of other traffic participants, camera data for a specific sensor, and detection and classification results. We highlight its performance in traffic and driving scenarios. We also present novel multi-agent simulation algorithms using reciprocal velocity obstacles that can model the behavior and trajectories of different traffic agents in dense scenarios, including cars, buses, bicycles and pedestrians. We also present novel methods for extracting trajectories from videos and use them for behavior modeling and safe navigation.

Bio: Dinesh Manocha is the Paul Chrisman Iribe Chair in Computer Science & Electrical and Computer Engineering at the University of Maryland College Park. He is also the Phi Delta Theta/Matthew Mason Distinguished Professor Emeritus of Computer Science at the University of North Carolina - Chapel Hill. He has won many awards, including Alfred P. Sloan Research Fellow, the NSF Career Award, the ONR Young Investigator Award, and the Hettleman Prize for scholarly achievement. His research interests include multi-agent simulation, virtual environments, physically-based modeling, and robotics. He has published more than 500 papers and supervised more than 36 PhD dissertations. He is an inventor of 10 patents, several of which have been licensed to industry. His work has been covered by the New York Times, NPR, Boston Globe, Washington Post, ZDNet, as well as DARPA Legacy Press Release. He was a co-founder of Impulsonic, a developer of physics-based audio simulation technologies, which was acquired by Valve Inc. He is a Fellow of AAAI, AAAS, ACM, and IEEE and also received the Distinguished Alumni Award from IIT Delhi. See http://www.cs.umd.edu/~dm

Gerard Guy Medioni

School of Engineering, University of Southern California

Title: Computer Vision Outside Academia: A Personal Journey

Abstract: I will present my personal perspective on the use of Computer Vision to solve real problems, with the goal to create products, not papers nor proof of concept demos. You will accompany me on my journey from the 80s to today, exploring several different applications, from halftone registration (Opti-copy) to virtual billboard insertion in live video (Symah Vision), face recognition (Geometrix), animated face swap and animation in video (Bigstage), 3D image capture and skeleton fitting (PrimeSense), and finally Amazon Go. From each of these, we’ll infer lessons about technology, business, and the tricky road to success or failure.

Bio: Professor Gérard Medioni received the Diplôme d’Ingenieur from ENST, Paris in 1977, a M.S. and Ph.D. from the University of Southern California in 1980 and 1983 respectively. He is currently on leave at Amazon as Director of Research for Amazon Go, from his position as Professor of Computer Science and Electrical Engineering, and co-director of the Institute for Robotics and Intelligent Systems (IRIS). He served as Chairman of the Computer Science Department from 2001 to 2007. Professor Medioni has made significant contributions to the field of computer vision. His research covers a broad spectrum of the field, such as edge detection, stereo and motion analysis, shape inference and description, and system integration. He has published 4 books, over 80 journal papers and 200 conference articles, and is the recipient of 19 international patents. He is the editor, with Sven Dickinson, of the Computer Vision series of books for Morgan-Claypool Prof. Medioni is on the advisory board of the IEEE Transactions on PAMI Journal, associate editor of the International Journal of Computer Vision, associate editor of the Pattern Recognition and Image Analysis Journal, associate editor of the International Journal of Image and Video Processing. He is vice president of the Computer Vision Foundation (CVF). Prof. Medioni served at program co-chair of the 1991 IEEE CVPR Conference in Hawaii, of the 1995 IEEE Symposium on Computer Vision in Miami, general co-chair of many CVPR Conferences (1997, 2001, 2007, 2009, 2020), conference co-chair of ICPR (1998, 2014), WACV (2009, 2011, 2013, 2015, 2017, 2019, 2021), ICCV (2017, 2019). Prof. Medioni has been a consultant to several companies and startups (DXO, Poseidon, Opti-copy, Geometrix, Symah Vision, KLA-Tencor, PrimeSense) prior to joining Amazon. He is a Fellow of IAPR, a Fellow of the IEEE, and a Fellow of AAAI.

Srinivasa Narasimhan

Robotics Institute, Carnegie Mellon University

Title: Computational Imaging and Illumination for Smart Devices

Abstract: In this talk, I will present several smart devices that are designed using a combination of programmable computational imaging and illumination. These include video-rate 3D sensors with the energy efficiencies of LIDARs, smart headlights that improve the driving experience in treacherous conditions and smart health devices that image deep below skin. These devices have strong impact on many application domains including autonomous driving, field robotics and health-care diagnosis.

Bio: Srinivasa Narasimhan is a Professor of Robotics and ECE (courtesy) at Carnegie Mellon University. He obtained his PhD from Columbia University in Dec 2003. His group focuses on novel techniques for imaging, illumination and light transport to enable applications in vision, graphics, robotics, agriculture and medical imaging. His works have received several awards: Best Demo Award (IEEE ICCP 2015), A9 Best Demo Award (IEEE CVPR 2015), Marr Prize Honorable Mention Award (2013), FORD URP Award (2013), Best Paper Runner up Prize (ACM I3D 2013), Best Paper Honorable Mention Award (IEEE ICCP 2012), Best Paper Award (IEEE PROCAMS 2009), the Okawa Research Grant (2009), the NSF CAREER Award (2007), Adobe Best Paper Award (IEEE Workshop on Physics based methods in computer vision, ICCV 2007) and IEEE Best Paper Honorable Mention Award (IEEE CVPR 2000). His research has been covered in popular press including NY Times, BBC, PC magazine and IEEE Spectrum and is highlighted by NSF and NAE. He is the co-inventor of programmable headlights, Aqualux 3D display, Assorted-pixels, Motion-aware cameras, Episcan3D, EpiToF3D, and programmable triangulation light curtains. He co-chaired the International Symposium on Volumetric Scattering in Vision and Graphics in 2007, the IEEE Workshop on Projector-Camera Systems (PROCAMS) in 2010, and the IEEE International Conference on Computational Photography (ICCP) in 2011, co-edited a special journal issue on Computational Photography, and serves on the editorial board of the International Journal of Computer Vision and as Area Chair of top computer vision conferences (CVPR, ICCV, ECCV, BMVC, ACCV).

Wei Qiao

Department of Electrical & Computer Engineering, University of Nebraska-Lincoln

Title: Electrical-Signature-Based Online Prognostic Health Monitoring for Wind Turbines

Abstract: Operation and maintenance (O&M) costs account for a significant portion (10%-35%) of the levelized cost of electricity generated by wind turbines. Currently in the wind power industry, maintenance is mainly performed on regular schedules or when significant damage occurs in a wind turbine making it inoperable, instead of being determined by the actual health condition of the wind turbine. To reduce the failure rate and level and O&M costs and improve the availability, reliability, safety, and lifespan of wind turbines, it is desirable to perform condition-based predictive maintenance for wind turbines, which will require a high-fidelity online health monitoring system. Dr. Qiao is the pioneer in the development of electrical-signature-based, low-cost, high-fidelity, online prognostic health monitoring techniques for wind turbines. His group has developed a suite of methods using electrical signals acquired from generator terminals or in the control systems of power electronic converters for online fault diagnosis and prognosis and remaining useful life prediction of wind turbines in all operating conditions. These techniques have been successfully applied to wind turbines of different types and different sizes from kW to MW scale. The work accomplished by Dr. Qiao provides a disruptive solution to address the challenges of high failure rates and high O&M costs facing the wind power industry. This seminar will present some of the pioneer work on electrical-signature-based online prognostic health monitoring for wind turbines led by Dr. Qiao.

Bio: Wei Qiao is a Professor with the Department of Electrical and Computer Engineering, the Director of the Power and Energy Systems Laboratory, and the Chair of the Electrical Engineering Graduate Program at the University of Nebraska-Lincoln, Lincoln, NE, USA. His research interests include renewable energy systems, magnetic devices, condition monitoring, power electronics, electric machines and drives, energy storage systems, power system control and optimization, and smart grids. He has published 230 peer-reviewed journal and conference proceeding papers, and holds 6 U.S. patents issued and 5 U.S./international patents pending. Dr. Qiao is an Editor of the IEEE Transactions on Energy Conversion and the IEEE Power Engineering Letters, and an Associate Editor of the IEEE Transactions on Power Electronics and the IEEE Journal of Emerging and Selected Topics in Power Electronics. He was an Associate Editor of the IEEE Transactions on Industry Applications and IET Power Electronics and a Corresponding Guest Editor of the IEEE Transactions on Industrial Electronics. Dr. Qiao was the recipient of a 2010 U.S. National Science Foundation CAREER Award and the 2010 IEEE Industry Applications Society Andrew W. Smith Outstanding Young Member Award. Professor Qiao received a Ph.D. degree from Georgia Institute of Technology, Atlanta, GA, USA, in 2008 and a master’s and bachelor’s degrees from Zhejiang University, Hangzhou, China, in 1997 and 2002, respectively, all in electrical engineering. He also received a master’s degree in High Performance Computation for Engineered Systems from Singapore-MIT Alliance (SMA) in 2003.

Tony Quek

Information Systems Technology and Design Pillar; Singapore University of Technology and Design

Title: AI: A Networking and Communication Perspective

Abstract: Recent breakthroughs in artificial intelligence and machine learning, including deep neural networks, the availability of powerful computing platforms and big data are providing us with technologies to perform tasks that once seemed impossible. In the near future, autonomous vehicles and drones, intelligent mobile networks, and intelligent internet-of-things (IoT) will become a norm. At the heart of this technological revolution, it is clear that we will need to have artificial intelligence over a massively scalable, ultra-high capacity, ultra-low latency, and dynamic new network infrastructure. In this talk, we will provide a simple overview of AI for the perspective of networking and communications and share some interesting applications. In addition, we will also share some of our preliminary works in this area.

Bio: Tony Q.S. Quek received the B.E. and M.E. degrees in Electrical and Electronics Engineering from Tokyo Institute of Technology, Tokyo, Japan, respectively. At Massachusetts Institute of Technology (MIT), Cambridge, MA, he earned the Ph.D. in Electrical Engineering and Computer Science. Currently, he is a tenured Associate Professor with the Singapore University of Technology and Design (SUTD). He also serves as the Acting Head of ISTD Pillar and the Deputy Director of SUTD-ZJU IDEA. His current research topics include wireless communications and networking, security, big data processing, network intelligence, and IoT. Dr. Quek has been actively involved in organizing and chairing sessions and has served as a TPC member in numerous international conferences. He is serving as the TPC Co-Chair for IEEE ISWCS 2019 and the General Chair for IEEE ICCC 2020. He is currently an elected member of the IEEE Signal Processing Society SPCOM Technical Committee. He was an Executive Editorial Committee Member of the IEEE Transactions on Wireless Communications, an Editor of the IEEE Transactions on Communications, and an Editor of the IEEE Wireless Communications Letters. He is a co-author of the book “Small Cell Networks: Deployment, PHY Techniques, and Resource Allocation” published by Cambridge University Press in 2013 and the book “Cloud Radio Access Networks: Principles, Technologies, and Applications” by Cambridge University Press in 2016. Dr. Quek received the 2008 Philip Yeo Prize for Outstanding Achievement in Research, the IEEE Globecom 2010 Best Paper Award, the 2012 IEEE William R. Bennett Prize, the 2016 IEEE Signal Processing Society Young Author Best Paper Award, 2017 CTTC Early Achievement Award, 2017 IEEE ComSoc AP Outstanding Paper Award, and 2016-2018 Clarivate Analytics Highly Cited Researcher. He is a Distinguished Lecturer of the IEEE Communications Society and a Fellow of IEEE.

Gaurav Sharma

Department of Electrical and Computer Engineering, University of Rochester

Title: Large Scale Visual Data Analytics for Geospatial Applications

Abstract: The widespread availability of high resolution aerial imagery covering wide geographical areas is spurring a revolution in large scale visual data analytics. Specifically, modern aerial wide area motion imagery (WAMI) platforms capture large high resolution at rates of 1-3 frames per second. The sequences of images, which individually span several square miles of ground area, represent rich spatio-termporal datasets that are key enablers for new applications. The effectiveness of such analytics can be enhanced by combining WAMI with alternative sources of rich geo-spatial information such as road maps or prior georegistered images. We present results from our recent research in this area covering three topics. First, we describe a novel method for pixel accurate, real-time registration of vector roadmaps to WAMI imagery based on moving vehicles in the scene. Next, we present a framework for tracking WAMI vehicles across multiple frames by using the registered roadmap and a new probabilistic framework that allows us to better estimate associations across multiple frames in a computationally tractable algorithm. Finally, in the third part, we highlight, how we can combine structure from motion and our proposed registration approach to obtain 3D georegistration for use in application such as change detection. We present results on multiple WAMI datasets, including nighttime infrared WAMI imagery, highlighting the effectiveness of the proposed methods through both visual and numerical comparisons.

Bio: Gaurav Sharma is a professor in the Departments of Electrical and Computer Engineering, Computer Science, and Biostatistics and Computational Biology, and a Distinguished Researcher in Center of Excellence in Data Science (CoE) at the Goergen Institute for Data Science at the University of Rochester. He received the PhD degree in Electrical and Computer engineering from North Carolina State University, Raleigh in 1996. From 1993 through 2003, he was with the Xerox Innovation group in Webster, NY, most recently in the position of Principal Scientist and Project Leader. His research interests include data analytics, cyber physical systems, signal and image processing, computer vision, and media security; areas in which he has 52 patents and has authored over 200 journal and conference publications. He currently serves as the Editor-in-Chief for the IEEE Transactions on Image Processing. From 2011 through 2015, he served as the Editor-in-Chief for the Journal of Electronic Imaging and, in the past, has served as an associate editor for the Journal of Electronic Imaging, the IEEE Transactions on Image Processing, and for the IEEE Transactions on Information Forensics and Security. He is a member of the IEEE Publications, Products, and Services Board (PSPB) and chaired the IEEE Conference Publications Committee in 2017-18. He is the editor of the Digital Color Imaging Handbook published by CRC press in 2003. Dr. Sharma is a fellow of the IEEE, a fellow of SPIE, a fellow of the Society for Imaging Science and Technology (IS&T) and has been elected to Sigma Xi, Phi Kappa Phi, and Pi Mu Epsilon. In recognition of his research contributions, he received an IEEE Region I technical innovation award in 2008.

Dinggang Shen

Center for Image Analysis and Informatics, UNC Chapel Hill, School of Medicine

Title: Deep Learning in Medical Image Synthesis and Its Applications

Abstract: This talk will introduce our five MICCAI 2018 papers on image synthesis with deep learning. Specifically, for reducing scanning time or potential risk, we proposed three single-modality based image synthesis methods. For example, to address the issue of missing modality such as PET in brain disease diagnosis, we developed 3D-CycleGAN for learning the bidirectional mappings between MRI and PET and thus imputing the missing PET for helping multimodality based disease diagnosis. Also, to enhance the quality of 3T MRI, we proposed a dual-domain cascaded regression for interactive learning at both spatial and frequency domains, with guidance from corresponding 7T MRI during the training. Also, by using the relationship between different sequences of images acquired in MR scanners as well as the nature of simultaneous acquisitions of PET and MR in PET/MR scanners, we proposed two multi-modality based image synthesis methods. For example, for fast T2 MRI acquisition, we developed a novel deep learning framework, based on dense Unet, to reconstruct T2 MRI from T1 MRI and under-sampled T2 MRI. To reduce dose for PET acquisition in PET/MR scanners, we developed locality-adaptive multi-modality GANs for estimating standard-dose PET from low-dose PET and MRI. The details of all these five methods will be provided in this talk, by also introducing both clinical significance and motivation for each developed method.

Bio: Dinggang Shen is Jeffrey Houpt Distinguished Investigator, and a Professor of Radiology, Biomedical Research Imaging Center (BRIC), Computer Science, and Biomedical Engineering in the University of North Carolina at Chapel Hill (UNC-CH). He is currently directing the Center for Image Analysis and Informatics, the Image Display, Enhancement, and Analysis (IDEA) Lab in the Department of Radiology, and also the medical image analysis core in the BRIC. He was a tenure-track assistant professor in the University of Pennsylvanian (UPenn), and a faculty member in the Johns Hopkins University. Dr. Shen’s research interests include medical image analysis, computer vision, and pattern recognition. He has published more than 1000 papers in the international journals and conference proceedings, with H-index 89. He serves as an editorial board member for eight international journals. He has also served in the Board of Directors, The Medical Image Computing and Computer Assisted Intervention (MICCAI) Society, in 2012-2015. He is General Chair for MICCAI 2019 in Shenzhen, China. He is Fellow of IEEE, Fellow of The American Institute for Medical and Biological Engineering (AIMBE), and Fellow of The International Association for Pattern Recognition (IAPR).

Li Shen

School of Computer, National University of Defense Technology

Title: Efficient Virtual Address Translation in GPUs

Abstract: Recent studies indicate that the impact of virtual-to-physical address translation on GPU performance is increasing, especially for irregular applications. This talk summarizes the ways to reduce address translation overheads. Firstly, we divide the process of address translation into several stages, and then introduce the optimization method of each stage. Finally, we will introduce a recently proposed method that can effectively reduce the cost of page table access.

Bio: Li Shen is a professor in Department of Computer Science at National University of Defense Technology. He received his PhD in Computer Science and Technology at National University of Defense Technology. His research interests lie in high performance processor architecture. Specific research topics include multi-core/many-core processor architecture, GPU architecture, performance and power consumption optimization, runtime environment, etc.

Zuoqiang Shi

Department of Mathematical Sciences Yau Mathematical Sciences Center, Tsinghua University

Title: PDE-Based Models in Machine Learning

Abstract: In this talk, I will present several PDE models and show their relations to machine learning and deep learning problem. In these PDE models, we use manifold to model the low dimensional structure hidden in high dimensional data and use PDEs to study the manifold. I will reveal the close connections between PDEs and deep neural networks. Theoretical analysis and numerical simulations show that PDEs provide us powerful tools to understand high dimensional data.

Bio: Sep. 2003 - Jul 2008: Ph.D in Applied Mathematics, Zhou Pei-yuan Center for Applied Mathematics, Tsinghua University, Beijing, China. Nov. 2006 - Feb. 2008: Special doctoral student in Applied & Computational Mathematics, Caltech, United States. Sep. 1999 - Jul 2003: B.S. in Mathematics, Department of Mathematical Science, Tsinghua University, Beijing, China. Position Sep. 2011 - : Associate professor, Yau Mathematical Sciences Center, Tsinghua University. Sep. 2008 - Sep. 2011 : Postdoctoral Scholar in Applied & Computational Mathematics, Caltech.

Jiwu Shu

Department of Computer Science and Technology, Tsinghua University

Title: Opportunities and Challenges of Persistent Memory to Computer Systems

Abstract: Data-driven processing has become the main processing mode in high-performance computing, big data analytics and artificial intelligence applications. The storage and processing capacity for data is critical to computer systems. Nowadays, storage systems face the strict requirements of new applications, at the same time, new storage hardware brings new opportunities for the design of storage systems. Persistent memory (such as 3D XPoint) connects to the CPU's persistent memory through a memory bus, providing memory-level data persistence. Persistence has changed the two-tier storage pattern in traditional storage systems. The boundaries of volatility and persistence have changed, and extremely low hardware latency imposes stringent requirements on software overhead. These changes have led to the changes in software hierarchy of persistent memory storage systems. This report will address the related issues, challenges and some research progress of the software layer in persistent memory storage systems. It will be introduced from three aspects: operating system, distributed system and new architecture to introduce the collaboration of user-mode and kernel-mode in storage system, distributed persistent memory systems and Processing in Memory (PIM).

Bio: Jiwu Shu is the professor of department of computer science and Technology, Tsinghua University. His research area includes: new storage systems and technologies based on non-volatile memory devices, solid-state storage systems and technologies based on flash devices, network/cloud/big data storage systems and key technologies, data storage reliability. His latest research direction includes: in-network/storage compute architect, storage system optimization for AI. He is the deputy director of Academic Work Committee in China Computer Federation, deputy director and Executive Member of Information Storage Technology Committee in China Computer Federation, deputy director of National Engineering Lab for Disaster Backup Technologies. He has won one second class of National Science and Technology Progress Award and one second class of National Science and Technology Invention Award. He is also an IEEE fellow, a CCF fellow.

Meixia Tao

Department of Electronic Engineering, Shanghai Jiao Tong University

Title: Computation Replication for Mobile Edge Computing

Abstract: Existing works on task offloading in mobile edge computing (MEC) networks often assume a task be executed once at a single edge node (EN). Downloading the computed result from the EN back to the mobile user thus may suffer long delay if the downlink channel experiences strong interference or deep fading. In this talk, we shall exploit the idea of computation replication in MEC networks to speed up the downloading phase. Computation replication allows each user to offload its task to multiple ENs for repetitive execution so as to create multiple copies of the computed result at different ENs which can then enable transmission cooperation and hence reduce the communication latency for result downloading. Yet, computation replication may also increase the communication latency for task uploading. The aim of this talk is to characterize asymptotically an order-optimal upload-download communication latency pair for a given computation load in a multi-user multi-server MEC network. Several insights on the fundamental computation-communication tradeoffs will be revealed.

Bio: Meixia Tao is a Professor with the Department of Electronic Engineering, Shanghai Jiao Tong University, China. She received the Ph.D. degree in electrical and electronic engineering from Hong Kong University of Science and Technology in 2003. Her current research interests include wireless caching, edge computing, physical layer multicasting, and resource allocation. Dr. Tao served as a member of the Executive Editorial Committee of the IEEE Transactions on Wireless Communications. She was on the Editorial Board of the IEEE Transactions on Wireless Communications (2007-2011), and the IEEE Transactions on Communications (2012-2018), the IEEE Communications Letters (2009-2012), and the \textsc{IEEE Wireless Communications Letters} (2011-2015). She serves as Symposium Oversight Chair of IEEE ICC 2019, Symposium Co-Chair of IEEE GLOBECOM 2018, the TPC chair of IEEE/CIC ICCC 2014 and Symposium Co-Chair of IEEE ICC 2015. Dr. Tao is a Fellow of IEEE. She is the recipient of the IEEE Marconi Prize Paper Award in 2019, the IEEE Heinrich Hertz Award for Best Communications Letters in 2013, the IEEE/CIC International Conference on Communications in China (ICCC) Best Paper Award in 2015, and the International Conference on Wireless Communications and Signal Processing (WCSP) Best Paper Award in 2012. She also receives the IEEE ComSoc Asia-Pacific Outstanding Young Researcher award in 2009.

C. K. Michael Tse

Department of Electronic and Information Engineering,Hong Kong Polytechnic University

Title: Modeling Cascading Failure in Power Systems from a Network Perspective

Abstract: Recent attempts in applying complex network analysis to the study of cascading failures have gained new insights into the effects of network structure on the extent and rapidity of failure events that occur in large-scale power systems. Results generated from such studies, though being able to shed some light on the effective assessment of the robustness of power systems, are not always consistent with historical data. Our recent efforts in incorporating physical power flow processes in the model of failure propagation and our latest findings using this new model have attracted much attention from the engineering community. The consistency with historical data verified the importance of incorporating physical processes in the model and the appropriate application of complex network concepts for the study of cascading failure in physical systems. This talk describes the recent progress in the study of cascading failures in physical systems.

Bio: C. K. Michael Tse obtained the BEng(Hons) and PhD degrees from the University of Melbourne, Australia, in 1987 and 1991, respectively. Presently, he is Chair Professor of Electronic Engineering with the Hong Kong Polytechnic University, and from 2005 to 2012 he was Head of its Department of Electronic and Information Engineering. His research interests include power electronics, nonlinear systems, and network applications. In 2005 he was elected IEEE Fellow. Prof. Tse received numerous research and invention prizes including IEEE Transactions Best Paper Prizes and Gold Medals at Geneva International Invention Exhibitions. He was conferred honorary professorships by a few Australian and Chinese universities and was awarded distinguished fellowships by a few Australian and Canadian universities. He serves/has served as Editor-in-Chief for IEEE Transactions on Circuits and Systems II, IEEE Circuits and Systems Magazine, as Editor for a few other journals, and as member of a few IEEE institute-level committees. Locally, he serves/has served on panels of ITF, RGC and ITC-ESS and on Quality Education Fund Committee of HKSAR Government, and is a broad member of the Hong Kong Sinfonietta.

Mengdi Wang

Department of Operations Research and Financial Engineering, Princeton University

Title: Reinforcement Leaning in Feature Space: Matrix Bandit, Kernels, and Regret Bound

Abstract: Exploration in reinforcement learning (RL) suffers from the curse of dimensionality when the state-action space is large. A common practice is to parameterize the high-dimensional value and policy functions using given features. However existing methods either have no theoretical guarantee or suffer a regret that is exponential in the planning horizon H. In this talk, we propose an online RL algorithm, namely the MatrixRL, that leverages ideas from linear bandit to learn a low-dimensional representation of the probability transition model while carefully balancing the exploitation-exploration tradeoff. We show that MatrixRL has an equivalent kernelized version, which is able to work with an arbitrary kernel Hilbert space without using explicit features. To our best knowledge, for RL using features or kernels, our results are the first regret bounds that are near-optimal in time T and dimension d and polynomial in the planning horizon.

Bio:

Qijie Wang

School of Electrical and Electronic Engineering, Nanyang Technological University

Title: Mid-infrared Photonics and Nanophotonics

Abstract: Mid-infrared (MIR) spectral region, hereafter defined as the ~3 – 20 μm wavelength range, hosts particular scientific and technological interests. Many molecules have strong and rich spectral fingerprints in this MIR region, therefore, MIR photonic and nanophotonic devices are potentially very promising for a breath of applications such as environmental and bio-chemical sensing, defense and security, industrial control, and medicine, etc. However, MIR remains a far from mature region, with a dearth in both passive and active components. The latter, including sources, modulators and photodetectors, often require materials with small enough direct bandgap, which, unfortunately, only exists in very few materials. In this talk, I am going to present the working principles and our achievements of MIR quantum cascade laser, which is a semiconductor based electrical pumping source that can be designed to cover the whole MIR regime. Then I am going to present our recent research in applying 2D materials for applications in the MIR regime, particularly room temperature broadband photodetectors [1-2]. In the end, I am also going to present how we achieve ultraconfined MIR surface phonon polariton devices in dichalcogenide 2D materials [3]. Reference [1] Xuechao Yu, Peng Yu, Di Wu, Bahadur Singh, Qingsheng Zeng, Hsin Lin, Wu Zhou, Zheng Liu and Qi Jie Wang, “Atomically-thin Noble Metal Dichalcogenide: A Broadband Mid-infrared Semiconductor”, Nature Communication, 9:1545, 2018. [2] Xuechao Yu, Yangyang Li, Xiaonan Hu, Daliang Zhang, Ye Tao, Md. Azimul Haque, Tom Wu, and Q. J. Wang, “Narrow bandgap oxide nanoparticles coupled with graphene for high performance mid-infrared photodetection”, 9, 4299, 2018. [3] A. M. Dubrovkin#, B. Qiang, H. N. S. Krishnamoorthy, N. I. Zheludev and Q. J. Wang, "Ultra-confined surface phonon polaritons inmolecular layers of van der Waals dielectrics", Nature Communications, 9, 1726, 2018.

Bio: Professor Wang Qijie is a full professor and the Associate Chair (Research) at School of Electrical and Electronic Engineering, Nanyang Technological University Singapore. He also holds a joint appointment at School of Physical and Mathematical Sciences at the same University. His research field is in the mid-infrared and Terahertz Photonics. He has published/co-published more than 140 papers in top international journals (like Science, Nature Photonics, Nature Materials, and Nature Communications). He was the recipient of the IES (Institution of Engineers Singapore) Prestigious Engineering Achievement Team Award of Singapore Twice in 2005 and 2017, respectively, 30th World Culture Special Recognition Award 2013, the prestigious Singapore Young Scientist Award 2014, and Nanyang Research Award NTU 2015.

Xunbin Wei

School of Biomedical Engineering, Shanghai Jiao Tong University

Title: From Optical Monitoring of Circulating Tumor Cells to Light Treatment of Alzheimer Disease

Abstract: Melanoma, developing from melanocytes, is the most serious type of skin cancer. Circulating melanoma cells, the prognosis marker for metastasis, are present in the circulation at the early stage. Thus, quantitative detection of rare circulating melanoma cells is essential for monitoring tumor metastasis and prognosis evaluation. Compared with in vitro assays, in vivo flow cytometry is able to identify circulating tumor cells without drawing blood. Here, we built in vivo photoacoustic flow cytometry based on the high absorption coefficient and monitoring circulating melanoma cells, which is useful for evaluating cancer stage and monitoring therapeutic effects. In addition, this talk will report the recent progress of our research work in the light treatment of Alzheimer disease.

Bio: Dr. Wei received his bachelor in physics from University of Science and Technology of China, Hefei. He received his PhD from Department of Physiology and Biophysics, University of California, Irvine. Dr. Wei completed his post-doc training at Children's Hospital, Harvard Medical School. From 2006-2010, he is a professor in Fudan University, China. Currently he is a professor and chair in Deparment of Biomedical Instrumentation, School of Biomedical Engineering, Shanghai Jiao Tong University, China. Dr. Wei is an SPIE Fellow, and recipient of Chinese Outstanding Young Scholar Award. He has published more than 80 peer-reviewed papers, including in Nature and PNAS. His research interests include cancer detection by optical means, optical manipulation of cells, and light treatment of Alzheimer disease.

Huaqiang Wu

Institute of Microelectronics, Tsinghua University

Title: Computation in Memristors: From Device to System

Abstract: Recently, computation in memory becomes very hot due to the urgent needs of high computing efficiency in artificial intelligence applications. In contrast to von-neumann architecture, computation in memory technology avoids the data movement between CPU/GPU and memory which could greatly reduce the power consumption. Memristor is one ideal device which could not only store information with multi-bits, but also conduct computing using ohm’s law. To make the best use of the memristor in neuromorphic systems, a memristor-friendly architecture and the software-hardware collaborative design methods are essential, and the key problem is how to utilize the memristor’s analog behavior. We have designed a generic memristor crossbar based architecture for convolutional neural networks and perceptrons, which take full consideration of the analog characteristics of memristors. Furthermore, we have proposed an online leanring algorithm for memristor based neuromorphic systems which overcomes the varation of memristor cells and endue the system the ability of reinforcement learning based on memristor’s analog behavior.

Bio: Dr. Huaqiang Wu is presently the deputy director of the Institute of Microelectronics, Tsinghua University, Beijing, China. Dr. Wu is also served as the director of Micro/Nano Fabrication Center of Tsinghua University and director of Beijing Innovation Center for Future Chips. Dr. Wu received his Ph.D. degree in electrical and computer engineering from Cornell University, Ithaca, NY, in 2005. Prior to that, he graduated from Tsinghua University, Beijing, China, in 2000 with double B.S. degrees in material science & engineering and enterprise management. From 2006 to 2008, he was a senior engineer and MTS in Spansion LLC, Sunnyvale, CA. He joined the Institute of Microelectronics, Tsinghua University in 2009. His research interests include emerging memory and neuromorphic computing technologies. Dr. Wu has published more than 100 technical papers and owns more than 90 US and China patents. Dr. Wu is the PI/Co-PI for more than 20 major national research projects. Dr. Wu has wide collaborations with companies worldwide including Samsung, SK Hynix, Applied Materials, Lam Research, Cisco, SMIC, Gigadevice, etc. Dr. Wu was the recipient of China Industry University Research Cooperation Innovation Award and Beijing Outstanding Young Talent Award.

Jing Xiang

Cincinnati Children's Hospital Medical Center, University of Cincinnati

Title: High Frequency Brain Signal – A New Frontier in Neuroimaging

Abstract: Recent success in detecting and localizing high-frequency brain signals (HFBS, > 70 Hz) with non-invasive magnetoencephalography (MEG) and electroencephalography (EEG) opens a new window for the study of the brain. HFBS are commonly referred as high frequency oscillations (HFOs), which include high gamma, ripples, fast ripples and very high frequency oscillations (VHFOs). One of the major applications of HFBS is to precisely localize functional brain areas in clinical practice to assess surgical resectability and reduce the possibility of functional deficit caused by operation. Another important application of HFBS is to surgically treat drug-resistant focal epilepsy. For many patients who have drug-resistant epilepsy, surgical resection of the region where seizure arise is often the only cure. However, the identification of this epileptogenic zone is often imprecise. Pre-surgical localization of epileptic HFBS may result in these patients with drug-resistant epilepsy being seizure free. In addition, HFBS also play important roles in many other disorders. For example, it has been shown that MEG HFBS can guide normalization of cortical excitability in migraine and reduce the incidence of headache attacks. Though HFBS are important new biomarkers and have important clinical applications, numerous challenges exist in the use of HFBS. First, analysis of HFBS requires a high sampling rate to digitize data. The size of high sampling rate data can be very large (big data). Big data pose a challenge for clinical data analysis. Second, given the massive amounts of high-sampling rate data, it is impractical to rely on conventional manual analysis of HFBS. Third, HFBS are typically much weaker than the conventional low frequency brains signals (< 70 Hz). Fourth, existing visual identification of HFBS is subjective, time consuming, and error prone. Therefore, it is essential to develop automatic and objective detection approaches to analyses of HFBS in MEG/EEG data. This presentation will introduce the state of the art technologies that can solve aforementioned problems and enable the development of intelligent neuroimaging to improve clinical outcomes.

Bio: Dr. Jing Xiang is the Director of Magnetoencephalography (MEG) Research at Cincinnati Children’s Hospital Medical Center. He is a Professor in the Departments of Pediatrics and Neurology at the University of Cincinnati. Dr. Xiang played a key role in building the world’s first pediatric MEG Laboratory at the Hospital for Sick Children, in Toronto, Canada. He also established the world’s first high frequency brain signal database (ClinicalTrials.gov Identifier: NCT00600717). He is a pioneer in clinical applications of high frequency brain signals and has published more than 160 peer-reviewed papers. Dr. Xiang received numerous awards for his outstanding contributions (e.g. Walter Berdon Award) and is an editor for several journals (e.g. Frontier in Human Neuroscience).

Yi Xing

Department of Pathology and Laboratory Medicine, University of Pennsylvania

Title: Elucidating Transcriptome Complexity Using RNA-seq Big Data and Deep Learning

Abstract: Human cells produce a large number of distinct mRNA and protein isoforms from individual gene loci via alternative processing and modifications of RNA. High-throughput RNA sequencing (RNA-seq) is a powerful tool for transcriptome-wide measurements of mRNA isoform complexity. Large consortium projects (ENCODE, Roadmap Epigenomics, GTEx, etc) have generated RNA-seq data on tens of thousands of samples along with a wide variety of other genomic and phenotypic measurements. These rich resources create unprecedented opportunities to study the variation, regulation, and functions of alternative isoforms. In this talk, I will describe our recent efforts in developing and applying big data and deep learning approaches for elucidating alternative splicing variation in human transcriptomes.

Bio: Dr. Yi Xing is the Francis West Lewis Chair and Founding Director of the Center for Computational and Genomic Medicine at the Children’s Hospital of Philadelphia (CHOP), and Professor of Pathology and Laboratory Medicine at the University of Pennsylvania (Penn). Prior to his appointment at CHOP and Penn, Dr. Xing was a Professor of Microbiology, Immunology, and Molecular Genetics at UCLA, and served as Program Director of UCLA’s Bioinformatics Interdepartmental Ph.D. Program. Dr. Xing received his B.S. in Molecular and Cellular Biology and B.E. in Computer Science and Technology from the University of Science and Technology of China (2001). He completed his Ph.D. training in Bioinformatics with Dr. Christopher Lee at UCLA (2001-2006), and his postdoctoral training with Drs. Wing Hung Wong and Matthew Scott at the Stanford University (2006-2007). Dr. Xing has an extensive publication record in bioinformatics, genomics, and RNA biology. His work has provided fundamental insights into the function, regulation, and evolution of post-transcriptional RNA processing in mammals. His current research merges the fields of computational biology, biomedical data science, RNA genomics, human genetics, precision medicine, and immuno-oncology.

Junjie Yao

Department of Biomedical Engineering, Duke University

Title: Breaking the Limits in Photoacoustic Imaging: Deeper, Smaller and More colorful

Abstract: By physically combining electromagnetic and ultrasonic waves, photoacoustic imaging (PAI) has proven powerful for multi-scale anatomical, functional, and molecular imaging. In PAI, a short-pulsed laser beam illuminates the biological tissue to generate a small but rapid temperature rise, which leads to emission of ultrasonic waves due to thermoelastic expansion. The high-frequency ultrasonic waves are detected outside the tissue by an ultrasonic transducer to form an image that maps the original optical energy deposition in the tissue. PAI seamlessly combines the rich optical absorption contrast of biological tissue with the high optically- or acoustically-determined spatial resolutions. My talk will focus on three major technical new fronts of PAI developed in our group. First, PAI has broken the penetration limit and achieved super-deep (~15 cm) imaging by using advanced internal light delivery, extending its applications into internal organ imaging on large animals and humans. Second, by using novel fabrication technologies in optics, acoustics and scanning, miniaturized photoacoustic microscopy has achieved handheld, wearable and head-mounted imaging of skin, brain, and organoids with high spatial–temporal resolutions. Third, taking advantage of a variety of newly developed near-infrared photoacoustic-specific contrast agents, PAI has achieved high-sensitivity high-specificity imaging of malignant cancer, tissue hypoxia, and neuronal activities.

Bio: Dr. Junjie Yao is currently Assistant Professor at the Department of Biomedical Engineering at Duke University, and a faculty member of Duke Center for In Vivo Microscopy, Duke Cancer Institute, Duke Institute of Brain Sciences, and Fitzpatrick Institute for Photonics. Dr. Yao received his B.E. (2006) and M.E. (2008) degrees in Biomedical Engineering from Tsinghua University, and his Ph.D. degree in Biomedical Engineering at Washington University (2013). Dr. Yao is the receipt of the IEEE Photonic Society Young Investigator Award. He serves on the editorial board in Scientific Reports, Quantitative Imaging in Medicine and Surgery, and Near-infrared and Laser Engineering. Dr. Yao’s research interest is in photoacoustic tomography (PAT) technologies in life sciences, especially in functional brain imaging and early cancer detection. Dr. Yao has published more than 100 articles in peer-reviewed journals such as Nature Methods, Nature Medicine, Nature Biomedical Engineering, Nature Communication, PNAS, and PRL. He (co-)invented photoacoustic Doppler-bandwidth flowmetry, photoacoustic oxygen metabolic microscopy, Super-resolution photoacoustic microscopy, fast-functional photoacoustic microscopy, and reversibly-switchable photoacoustic tomography. Dr. Yao’s lab (PI-Lab) centers on developing beak-through PAT technologies with novel and advanced imaging performance, in terms of spatial resolutions, imaging speed, penetration depth, detection sensitivity, and functionality. Dr. Yao’s lab is interested with all aspects of PAT technology innovations, including efficient light illumination, high-sensitivity ultrasonic detection, super-resolution PAT, high-speed imaging acquisition, novel PA genetic contrast, and precise image reconstruction. On top of the technological advancements, Dr. Yao’s lab is devoted to serve the broad life science and medical communities with matching PAT systems for various research and clinical needs, especially for studying tumor angiogenesis, cancer hypoxia, and brain disorders. More research at http://photoacoustics.pratt.duke.edu/.

Nicholas Zabaras

College of Engineering, University of Notre Dame

Title: Physics-Informed LearnIng for Multiscale Systems

Abstract: Surrogate modeling and uncertainty quantification tasks for systems governed by PDEs are most often considered as supervised learning problems where input and output data pairs are used for training. The construction of such emulators is by definition a Small Data problem which poses challenges to modern tools such as Deep Learning approaches that have been developed to operate in a Big Data regime. Even in cases where such models have been shown to have good predictive capability in high-input dimensions, they fail to address constraints in the data implied by the PDE model. We will present a methodology in this direction that incorporates the governing equations of the physical model in the loss/likelihood functions. The resulting physics-constrained, deep learning models operate without labeled data (i.e. employing only training input data) and provide predictive responses that obey the constraints of the problem at hand just as well as typical deterministic models. This work employs a convolutional encoder-decoder convolutional neural network approach as well as a conditional flow-based generative model for the solution of PDEs, surrogate model construction for PDEs, and uncertainty quantification tasks. We will highlight how these techniques can be extended to learn Dynamics of complex systems and how using Geometric Convolutions can be enhanced to predict solutions of multiscale PDEs on optimized unstructured geometries. The above methods go beyond PDE systems and are directly applicable to discrete atomistic models in Chemistry, Biology and elsewhere. Wea data-free approach to coarse-graining in atomistic simulations using generative models based on variational auto-encoders.

Bio: Prof. Nicholas Zabaras joined Notre Dame in 2016 as the Viola D. Hank Professor of Computational Science and Engineering. He is the Director of the interdisciplinary University of Notre Dame “Center for Informatics and Computational Science (CICS)” that aims to bridge the areas of data-sciences, scientific computing and uncertainty quantification for complex multiscale/multiphysics problems in science and engineering. Among the various appointments, Prof. Zabaras was until recently the Hans Fisher Senior Fellow with the Institute for Advanced Study at the Technical University of Munich where he currently holds the position of "TUM Ambassador". He served for nearly 23 years at all academic ranks on the faculty of Cornell University.

Pinjia Zhang

Department of Electrical Engineering, Tsinghua University

Title: The Theory and Method of Non-Intrusive On-Line Monitoring of Electrical Equipment Based on Leakage Current Measuring

Abstract: Electrical equipment plays an important role in modern industrial applications. With the rapid development of electrification, the reliability of electrical equipment becomes more and more important. On-line monitoring can realize continuous condition monitoring of electrical equipment, provide early warning before the failure occurs, and avoid cascading reactions and greater economic losses caused by sudden failure. Non-intrusive monitoring method based on leakage current measuring will not have any impact on the normal operation of the original equipment, which provides great potential. This report discusses the theory and method of leakage current-based on-line monitoring of electrical equipment, including motors, transformers and cables. The ageing and failure mechanism of the electrical equipment will be discussed, together with the monitoring scheme and data processing. Taking the past research experience of the team in this area as an example, the report will show the technology of non-intrusive on-line monitoring of electrical equipment based on leakage current and the trend of electrical equipment health management.

Bio: Pinjia Zhang, associate professor, the Department of Electrical Engineering, Tsinghua University, senior member of the IEEE. He obtained his PhD from Georgia Tech, Atlanta, GA, USA in 2010. He was with the electric machines lab, GE Global Research from 2010 to 2015. He is a member of the Standard Committee and Award Committee of IEEE Industrial Application Society, Secretary-General of Beijing Branch of IEEE Industrial Application Society. He serves as the associate editor IEEE Transactions on Industrial Electronics, IEEE Transactions on Industry Applications, and IEEE Access. He also serves as the convener of CIGRE/A1.45 on-line monitoring standards committee for large generator systems. His research mainly focuses on the on-line monitoring and fault prognosis of electrical equipment. He has received 4 best paper awards by IEEE Industrial Application Society and Industrial Electronics Society. He has published over than 80 papers in refereed conference proceedings and journals, and has more than 40 patents granted. He is also the recipient of the 2018 IAS Andrew W. Smith Outstanding Young Member Achievement Award.

Tao Zhou

AMSS, Chinese Academy of Sciences

Title: Forward and Inverse Uncertainty Quantification for PDEs

Abstract: This talk is concerned with uncertainty quantification for PDEs. For the forward model, we shall review recent progresses on the analysis results for polynomial based approximations. For the inverse problems, the polynomial approximation is widely used as a surrogate model in the Bayesian inference to speed up the Markov chain Monte Carlo calculations. However, the use of such a surrogate introduces modeling errors that may severely distort the estimate of the posterior distribution. We thus present an adaptive procedure to construct a multi-fidelity polynomial surrogate. The key idea is to construct and refine the multi-fidelity surrogate over a sequence of samples adaptively determined from data so that the approximation can eventually concentrate to the posterior distribution. We also introduce a multi-fidelity surrogate based on the deep neural networks to deal with problems with high dimensional parameters.

Bio: Tao Zhou is currently an Associate Professor in Chinese Academy of Sciences (Beijing). Dr. Zhou’s research interests include Uncertainty Quantification, Phase Field Models, Backward SDEs, and Stochastic Optimal Control He serves as an associate editor for International Journal for UQ (IJUQ) and communications in Computational Physics (CiCP). Since 2019, he has been the Managing Editor for East Asian journal for Applied Mathematics. Dr. Zhou was a recipient of the NSFC Career Award for Excellent Young Scholars and CSIAM Excellent Young Scholar Prize.

Confirmed Speakers

Bo An

Xiang Bai

Enhong Chen

Xudong Chen

Frank Dellaert

Atilla Eryilmaz

James Gee

Richard Hartley

Thomas Hou

Jianwei Huang

Longbo Huang

Sheng-Kwang Hwang

Nan Jiang

Tao Jiang

Xiangyan Kong

Puxiang Lai

Lawrence Le

Xiuling Li

Yixue Li

Ming Lin

Jia Liu

Jiaying Liu

Ling Liu

Si Liu

John C.S. Lui

Patrick Luk

Ke Ma

Dinesh Manocha

Gerard Guy Medioni

Srinivasa Narasimhan

Wei Qiao

Tony Quek

Gaurav Sharma

Dinggang Shen

Li Shen

Zuoqiang Shi

Jiwu Shu

Meixia Tao

C. K. Michael Tse

Mengdi Wang

Qijie Wang

Xunbin Wei

Huaqiang Wu

Jing Xiang

Yi Xing

Junjie Yao

Nicholas Zabaras

Pinjia Zhang

Tao Zhou

Speakers and Speeches Information

Title: When AI Meets Game Theory

Title: OCR in the Wild: Current Situation, Challenges and Future Trends

Title: Exploiting Knowledge Tracing for Intelligent Education

Title: Solving Electrical Impedance Tomography via Deep Learning Approach

Title: Factor Graphs for Flexible Inference in Robotics and Vision

Title: Leveraging Side-Information for Learning and Optimization under Uncertainty with Applications in Social and Communication Networks

Title: Radiological Differential Diagnoses by Machine

Title: Mathematics and Geometry, Their Role in Computer Vision in the Deep Learning Age

Title: Solving Multiscale Problems by Integrating Physical Models with Deep Learning

Title: Economics of Mobile Crowd Sensing

Title: Multi-armed Bandits with Compensation

Title: Photonic Microwave Processing Using Nonlinear Dynamics of Semiconductor Lasers for Radio-over-fiber Applications

Title: Sample-efficient Exploration in Reinforcement Learning with Function Approximation

Title: Predicting Isoform Functions from Sequences and Expression Profiles via Deep Learning

Title: Improvement of 36-channel Magnetocardiography System Based on DC SQUID Magemetometers

Title: Wavefront Shaping-enabled Optical Focusing and Its Application Towards Deep-tissue Single Neuron Stimulation

Title: Ultrasonic Imaging of Cortical Bones

Title: Extreme Miniaturization of Passive Electronic Components by Self-Rolled-up Membrane Nanotechnology

Title: Genomics study of animal models

Title: Reconstructing Reality: From Physical World to Virtual Environments

Title: Combinatorial Bandits with Fairness Constraints

Title: Intelligent Visual Computing

Title: Robustness of Deep Learning Systems against Adversarial Inputs

Title: Human-centric Image Analysis

Title: An Online Learning Approach to Network Application Optimization with Guarantee

Title: Towards a New Energy Eco-system: The Convergence of the Electric Grid and The Electric Vehicle

Title: Mission Profile Based Reliability Analysis of Power Electronics System

Title: Autonomous Driving: Simulation and Navigation

Title: Computer Vision Outside Academia: A Personal Journey