HIT SEIE Academic Activities : Invited Talk by Prof. Simon Shaolei Du

Harbin Institute of Technology Advanced Communications Technologies Forum 2019

Title:

Understanding Optimization and Generalization in Deep Learning: A Trajectory-based Analysis

Date and Time: 9:00-11:00, January 4th, 2019

Location:ROOM 1011, BUILDING 2A, NO.2 YIKUANG STREET,

HARBIN, HEILONGJIANG, CHINA

Abstract:

This presentation will present recent progress on understanding deep neural networks by analyzing the trajectory of the gradient descent algorithm. Using this analysis technique, we are able to explain:

1) Why gradient descent finds a global minimum of the training loss even though the objective function is highly non-convex, and

2) Why a neural network can generalize even the number of parameters in the neural network is more than the number of training data.

Biography:

Simon Shaolei Du , Ph.D Department of Computer Science, Carnegie Mellon University, whose tutors are Prof.  Aarti Singh and Prof. Barnabás Póczos. His research interests include theoretical machine learning and statistical topics such as in-depth learning, matrix decomposition, convex/nonconvex optimization, transfer learning, reinforcement learning, nonparametric statistics, and robust statistics. In 2015, he received a double degree in Electrical Engineering and Computer, Engineering Mathematics and Statistics from the University of California at Berkeley. He is a Berkeley EECS winner, a member of Etta Kappa Nu, and a member of Phi Beta Kappa. As a leading young scholar in international machine learning, he is the 2018 NeurIPS top conference Best paper winner, the 2018 NeurIPS top conference NVIDIA Pioneer Award winner, and 19 computer top conference papers. He has also worked in research labs at Microsoft and Facebook.

 

 

HIT SEIE Academic Activities : Invited Talk by Prof. Tad Matsumoto

Harbin Institute of Technology Advanced Communications Technologies Forum 2018

Title 1: Introduction to Frequency Domain Turbo Equalization and its Applications

Date and Time: 9:00-11:30, September 22, 2018

Location:ROOM 1011, BUILDING 2A, NO.2 YIKUANG STREET,

HARBIN, HEILONGJIANG, CHINA

Abstract:

The goal of this lecture is to provide the audience with understanding of “turbo principle”.  To achieve this goal, this lecture will be started with the sliding window soft cancellation minimum mean squared error filtering (SC-MMSE) turbo equalization for single carrier signal transmission over frequency selective fading channels with relatively short memory. Then, this lecture introduces time-domain block-wise processing, and shows that no gains achieved by making modifications from sliding window to block wise processing due to its intractably heavy computational complexity.  Plus, this lecture further modifies the time-domain block wise SC-MMSE turbo equalization to frequency domain (FD SC-MMSE), where it is shown that the required computational complexity is very light and is constant regardless of the channel memory length.

Also, this lecture provides introductory explanations to the extrinsic information transfer (EXIT) chart as a tool for evaluating the efficiency of mutual information exchange. Information theoretic convergence analysis of the FD-SC MMSE turbo equalization, exemplifying the turbo principle, are provided in this lecture Furthermore, the concept of the FD SC-MMSE turbo equalization is applied to multiple input multiple output (MIMO) systems as a reasonable extension of the technique.  It is shown that when analyzing the mutual information exchange for MIMO turbo equalization, multi-dimensional EXIT plane has to be used.  Finally, trends and future prospects of research work towards broadband mobile communication systems are introduced, including turbo equalization of orthogonal frequency division multiplexing (OFDM) and single carrier (SC)-FDMA as well as the technique for eliminating the cyclic prefix (CP) from block-wise processing based FD SC-MMSE turbo equalization.

 

Title 2: Turbo Equalization and its Information Theoretic Analysis

Date and Time: 18:00-20:30, September 25, 2018

Location:ROOM 1011, BUILDING 2A, NO.2 YIKUANG STREET,

HARBIN, HEILONGJIANG, CHINA

Abstract:

A goal of this talk is to provide audience with the knowledge about the relationship between relay systems and the Distributed Coding techniques for correlated sources. To achieve this goal, this lecture is started by the re-enforcement of understanding of turbo principle, especially, frequency-domain soft cancellation minimum mean square error (FD SC-MMSE) based multiple-input multiple-output (MIMO) turbo equalization (This part is provided in Lecture-1). This lecture uses a lot of multi-dimensional extrinsic information transfer (EXIT) analysis to reveal the convergence properties of the FD SC-MMSE MIMO equalization, and identifies the optimal, close capacity achieving structure. It is shown that even with very simple serially concatenated convolution code with the component codes being very simple memory one codes can achieve near-capacity performance. Furthermore, the inner code, which is a very simple memory one recursive code, can eliminate the error floor due to the intersection of the EXIT curves, resulting in very sharp shape of the turbo cliff. This lecture also makes comparison of the shape of the EXIT curves with recursive and non-recursive convolutional codes.

At the final part of this lecture, we intentional “add” binary errors randomly between the MIMO antennas, and analyses the impacts of the “artificial errors”. It is shown that if the FD SC MMSE equalizer can utilize the error probability to modify the log likelihood ratio (LLR) in the vertical iteration, we can eliminate the effect of the “artificial errors”. The “artificial error” probability can be estimated only at the decoder side (no side information needed).

Obviously the “artificial errors” inserted in the connections between the antennas correspond to the “intra-link” errors in distributed lossy forwarding cooperative networks. Therefore, the last part of this lecture is a preparation of Lecture 3, Tutorial on Lossy Forward Relaying: Orthogonal and non-Orthogonal cases.

Biography:

Prof. Matsumoto received his bachelor’s and master’s degrees in electrical engineering from Keio University in Japan in 1978 and 1980, and his Ph.D. in 1991. He has taught at York University in Canada, Ilmenau University of Technology, Germany, and Keio University in Japan. Since April 2002, Prof. Matsumoto has taught at the University of Oulu, Finland. He has taught MIMO communication technology and turbo code principles and applications. Since 2007, he has also taught at the University of Japan’s Hokuriku Science and Technology University. , teaching two courses of information theory and coding theory.

Prof. Matsumoto’s research direction is mainly information theory and coding, especially the coding and equalization techniques of turbo codes. Prof. Matsumoto has been involved in the field of turbo code for more than 15 years. In 2004, he discussed the application of turbo code correlation technology to receivers for wireless communication, and then discussed extending the application of turbo codes to performance analysis in communication scenarios such as multi-antenna. The above research is included in Mobile Broadband Multimedia. In Networks Techniques, Models and Tools for 4G.

Prof. Matsumoto also studied the communication performance improvement of the iterative process for cooperative communication methods, and has published articles in a large number of well-known journals. In addition, he has served as an editor of several international academic journals. Prof. Matsumoto has received IEEE Outstanding Contribution Award, IEICE Best Paper, IEEE Outstanding Review, IEEE Distinguished Speaker, Finnish Distinguished Professor and other awards and titles since 2002. . At the same time, he is also IEEE Fellow, which has a great influence in the field of communication.

 

HIT SEIE Academic Activities : Invited Talk by Prof. Yungh-Siang Han

Harbin Institute of Technology Advanced Communications Technologies Forum 2018

Title 1: Efficient Decoding over Unknown Impulsive Noise Channels

Date and Time: 10:00-11:30, September 2, 2018

Location:ROOM 1011, BUILDING 2A, NO.2 YIKUANG STREET,

HARBIN, HEILONGJIANG, CHINA

Abstract:

It has been known from many researches that communication systems are susceptible to memoryless impulsive noise characterized by, for instance, the Bernoulli-Gaussian model. In order to combat this obstacle, channel coding has long served as an effective tool, especially in the context of moderately frequent occurrence of impulses, when the statistics of impulsive noise can be realized at the decoder. In this talk, irrespective of the statistics of impulses, an efficient decoding scheme is introduced by incorporating clipping-featured technique into the Viterbi algorithm. As a result, the proposed decoding scheme, while having a complexity at the same order as that of the Viterbi algorithm, is on a par with its optimal counterpart, for which statistics of impulses is assumed known at receiver, in terms of bit error probability. In addition, the Chernoff bounds of the bit error probabilities of the devised decoding algorithm are derived for both Bernoulli-Gaussian noise model and Middleton Class-A noise model. Comparisons between the bounds we derived and the simulated error rates under a variety of settings indicate that the ensuing analysis can provide critical insights for the efficacy of the proposed decoding approach when dealing with precarious frequent strong impulses.

 

Title 2: Novel FFT over Binary Finite Fields and Its Application to Reed-Solomon Erasure Codes

Date and Time: 10:00-11:30, September 7, 2018

Location:ROOM 1011, BUILDING 2A, NO.2 YIKUANG STREET,

HARBIN, HEILONGJIANG, CHINA

Abstract:

Abstract A fundamental issue in algebra is to reduce the computational complexities of arithmetic operations over polynomials. Many fast polynomialrelated algorithms, such as encoding/decoding of Reed-Solomon codes, are based on fast Fourier transforms (FFT). However, it is algorithmically harder as the traditional fast Fourier transform (FFT) cannot be applied directly over characteristic-2 finite fields. To the best of our knowledge, no existing algorithm for characteristic-2 finite field FFT/polynomial multiplication has provably achieved O(h log2(h)) operations. In this talk, we present a new basis of polynomial over finite fields of characteristic-2 and then apply it to the encoding/decoding of Reed-Solomon erasure codes. The proposed polynomial basis allows that h-point polynomial evaluation can be computed in O(h log2(h)) finite field operations with small leading constant. As compared with the canonical polynomial basis, the proposed basis improves the arithmetic complexity of addition, multiplication, and the determination of polynomial degree from O(h log2(h) log2 log2(h)) to O(h log2(h)). Based on this basis, we then develop the encoding and erasure decoding algorithms for the (n = 2r; k) Reed-Solomon codes. Thanks to the efficiency of transform based on the polynomial basis, the encoding can be completed in O(n log2(k)) finite field operations, and the erasure decoding in O(n log2(n)) finite field operations. To the best of our knowledge, this is the first approach supporting Reed-Solomon erasure codes over characteristic-2 finite fields while achieving a complexity of O(n log2(n)), in both additive and multiplicative complexities. As the complexity of leading factor is small, the algorithms are advantageous in practical applications.

This work was presented at the 55th Annual Symposium on Foundations of Computer Science (FOCS 2014).

Biography:

Prof. Yungh-Siang Han graduated from the Department of Electrical Engineering of Tsinghua University in Taiwan in 1984 and obtained a master’s degree in the same department in 1986. In 1993, Prof. Yungh-Siang Han received his Ph.D. in Computer and Information Science from Syracuse University, New York. He has taught at Hua Fan College of Humanities and Technology, Jinan International University, and Taipei University. From August 2010 to January 2017, he taught at the Department of Electrical Engineering of the Taiwan University of Science and Technology and was appointed as a lecturer in the school in June 2011. Since February 2015, he is also a lecturer at Taipei University. He is currently a Distinguished Professor of Distinguished Talents in the School of Electrical Engineering and Intelligence at Dongguan Institute of Technology.

Prof. Yungh-Siang Han’s research interests are mainly in error control codes, wireless networks and information security. He has been engaged in the most advanced error control code decoding research for over 20 years. Twenty years ago he first developed a continuous decoding algorithm based on the A* algorithm. At the time, the algorithm attracted a lot of attention because it was the most efficient maximum likelihood softness decision decoding algorithm for binary linear block codes. This decoding algorithm has been included in the classic textbook of error control codes.

Prof. Yungh-Siang Han also successfully applied coding theory to the research field of wireless sensor networks. He has published several highly cited works on wireless sensor network research. One of the random key pre-allocation schemes was cited more than two thousand times. He is also the editor of several international academic journals. Professor Han is the winner of the 1994 Ph.D. Thesis at Syracuse University and an IEEE Fellow. In 2013, one of his papers won the prestigious ACM CCS Test of Time award. This award is the most influential paper award of the year in ACM’s information security field.

IEEE ComSoc Harbin Chapter Members Reach the New Milestone

IEEE ComSoc Harbin Chapter Members Reach the New Milestone

In order to attract more talents and enhance the international influence, IEEE ComSoc Harbin Chapter organized a series of recruiting events facing to faculties and students of northeastern universities in China since 2018.

The core volunteers had made a lot of effort to promote IEEE and ComSoc. Chapter Chair, Prof. Weixiao Meng and Vice Chair, Dr. Shuai Han organized a publicity meeting where they described the history, development and current influence of the IEEE ComSoc Harbin Chapter. Faculty members also share the functions of IEEE ComSoc with their students and encourage them to start their professional careers with ComSoc student member. At the same time, Harbin Chapter also made full use of multimedia social networks. After contacting with professors who are focusing on information and communications engineering in northeastern universities, QQ and WeChat membership groups were established among universities and various organizations, mainly oriented to teachers, master and Ph.D students. In those groups, chapter issued the relevant notifications for the recruiting events and sent warm invitations to the potential teachers and students. In order to organize some activities afterward, Harbin Chapter raised funds to customize T-shirts for the new members and guests.

Due to meticulously preparation, many teachers and students responded positively. The number of active members in IEEE ComSoc Harbin Chapter has reached 121 by April, 2018. According to IEEE Current Grade Description, there are 1 associate member, 38 graduate student members, 51 members, 12 senior members, 17 student members and 2 affiliate. Among the 121 members, 64 of them are new members from last September, and more than 50% of new members are student or graduate student members.

Through data analysis, we can find that IEEE ComSoc Harbin Chapter membership is the absolute majority in Harbin Section. They mainly focus on computer sciences and information, engineering and technical communication.

Besides, the Harbin Chapter nominated Ms. Qian Chen, a Ph.D student member as the student chair, who plays an important bridge role between chapter and all kinds of students.

In the future, IEEE ComSoc Harbin Chapter will organize a series of lectures and academic activities for members in different areas to boost communication and cooperation. It is believed that with the joint efforts of all members, Harbin chapter will develop in a better way.

IEEE ComSoc Harbin Chapter Won Three International Academic Activities Awards at IEEE Globecom 2018

IEEE ComSoc Harbin Chapter Won Three International Academic Activities Awards at IEEE Globecom 2018

IEEE ComSoc Harbin Chapter won three international academic activities awards at IEEE Globecom 2018 in Abu Dhabi, United Arab Emirates from December 9th to13th in 2018, which are 2018 Chapter Of The Year Award, Asia Pacific Region 2018 Chapter Achievement Award, 2018 IEEE Asia Pacific Conference Contribution Award, and the chair of Harbin Chapter, Prof. Weixiao Meng, won 2018 Member and Global Activites Contribution Award.

IEEE Communications Society has 4 regions and 25,600 members currently in the world. China belongs to the Asia-Pacific, which includes 6 chapters.As the youngest one, IEEE ComSoc Harbin Chapter contains 147 members from universities and institutions in three provinces of Northeast China. The chapter is a non-profit academic organization aiming to lead the professional career of members, provide academic exchange platforms and service channels for members so as to promote the development of communication theory, technology and industry.

Harbin Chapter was founded by Prof. Weixiao Meng in 2012, who was also the chair. Associate Prof. Shuai Han served as the vice chair and established the student chair. In the past 6 years, Harbin Chapter has made outstanding achievements in academic exchanges, membership services, upgrades, development, information consultation and won consistent commendation from IEEE Communications Society. This is the first time for Chinese academic organizations to win the serialization of international academic activities award. Prof. Weixiao Meng attended the conference,accepted the award from the international academic organization and made a speech.

In the past one year, IEEE ComSoc Harbin Chapter tried its best to attract more elites and promote quantity and quality of its membership. In order to enrich the chapter   and elevate its international influence, Harbin Chapter organized a series of recruiting events facing to faculties and students of northeastern universities in China which has relative great influence.

The core volunteers has made a lot of effort to enhance the IEEE and ComSoc. Chapter Chair, Prof. Weixiao Meng and Vice Chair, Dr. Shuai Han organized a publicity meeting where they described the history, development and current influence of the IEEE ComSoc Harbin Chapter in details, which have become one of routine activities for newcomers every September. Prof. Weixiao Meng kept contacting with all the EEE and ECE Deans of universities from three provinces of Northeast China, encouraged them to distribute the recruiting polices to the corresponding areas of their own universities. Faculty members also shared the functions of IEEE ComSoc with their students and inspired them to start their professional careers with ComSoc student members. At the same time, Harbin Chapter noticed the crutial state of multimedia, so it made full use of social networks for more sufficient propaganda. After contacting with professors who are focusing on information and communications engineering in northeastern universities, QQ ,WeChat and other social communication groups which mainly oriented to teachers, master and Ph.D students were established among universities and various organizations. IEEE ComSoc Harbin Chapter nominated Ms. Qian Chen, a Ph.D student member as the student chair, who plays an important bridge role between chapter and all kinds of students in March, 2018.In those groups, Harbin Chapter issued the relevant notifications for the recruiting events and sent warm invitations to the potential teachers and students. In order to organize some activities afterward convieniently, Harbin Chapter raised funds to customize T-shirts for the new members and guests, and mailed them to everyone, which were extremely popular among them. Since most of students had no VISA/MASTER cards, the payment of membership became knotty. A faculty member was willing to pay equivalent dollars online for new student members so that the problem was succesfully solved. From these details, therefore, newcomers increased their confidence to Harbin Chapter.

Due to meticulously preparation, many teachers and students responded positively. In order to remain its lasting competitiveness, Harbin Chapter held more than 10 visiting academic lectures last year and hosted 3 DLT and DLS in recent years. In addition, they drafted newsletters to ComSoc twice a year. Their Monday Seminar is still ongoing chairing by student members in Harbin once a week.

Thanks to IEEE Communications Society for confirming the achievements of IEEE Comsoc Harbin Chapter in the last year and awarding the three awards for it. Harbin Chapter will keep seriousness, pragmatic and innovation, try to make greater effort for advanced contributions to IEEE Communications Society in the future.

HIT SEIE Summer Academic Activities VI: Invited Talk by Prof. Lingyang Song

Harbin Institute of Technology Advanced Communications Technologies Forum 2017

Title:Game theory for Big Data Processing

Date and Time: 11:30—13:00, Auguest 8, 2017

Location:ROOM 1013, BUILDING 2A, NO.2 YIKUANG STREET,

HARBIN, HEILONGJIANG, CHINA

Abstract: Modern communication networks are becoming highly virtualized with two-layer hierarchies, in which controllers at the upper layer with tasks to achieve ask a large number of agents at the lower layer to help realize big data processing. In this talk, we combine optimization approaches and game theory to address the tradeoff and convergence issues for incentive mechanism design in hierarchies. Specifically, we propose a multiple-leader multiple-follower (MLMF) game-based alternating direction method of multipliers (ADMM) that incentivizes the agents to perform the controllers’ tasks in order to satisfy the corresponding objectives for both controllers and agents. Both analytical and simulation results verify that the proposed method reaches a hierarchical social optimum and converges at a linear speed. More importantly, the convergence rate is independent of the network size, which indicates that the MLMF game-based ADMM can be used in a network with a very large size for big data processing.

Biography:

Lingyang Song (S’03-M’06-SM’12) received his PhD from the University of York, UK, in 2007, where he received the K. M. Stott Prize for excellent research. He worked as a research fellow at the University of Oslo, Norway until rejoining Philips Research UK in March 2008. In May 2009, he joined the School of Electronics Engineering and Computer Science, Peking University, China, as a full professor. His main research interests include MIMO, cognitive and cooperative communications, security, and big data. Dr. Song wrote 2 text books, “Wireless Device-to-Device Communications and Networks” and “Full-Duplex Communications and Networks” published by Cambridge University Press, UK. He is the recipient of IEEE Leonard G. Abraham Prize in 2016 and IEEE Asia Pacific (AP) Young Researcher Award in 2012. He is currently on the Editorial Board of IEEE Transactions on Wireless Communications. He is an IEEE distinguished lecturer since 2015.

 

HIT SEIE Summer Academic Activities V: Invited Talk by Prof. Nei Kato

Harbin Institute of Technology Advanced Communications Technologies Forum 2017

Title:The Deep Learning Vision for Heterogeneous Network Traffic Control: Proposal, Challenges, and Future Perspective

Date and Time: 10:00—11:30, Auguest 8, 2017

Location:ROOM 1013, BUILDING 2A, NO.2 YIKUANG STREET,

HARBIN, HEILONGJIANG, CHINA

Abstract:

Recently, deep learning, an emerging machine learning technique, is garnering a lot of research attention in several computer science areas. However, to the best of our knowledge, its application to improve heterogeneous network traffic control which is an important and challenging area for IoT by its own merit has yet to appear because of the difficult challenge in characterizing the appropriate input and output patterns for a deep learning system to correctly reflect the highly dynamic nature of large-scale heterogeneous networks. In this talk, an appropriate input and output characterizations of heterogeneous network traffic will be introduced and a supervised deep neural network system will be proposed. I will describe how our proposed system works and how it differs from traditional neural networks. Also, preliminary results will be discussed and I will demonstrate the encouraging performance of our proposed deep learning system compared to a benchmark routing strategy (Open Shortest Path First (OSPF)) in terms of significantly better signaling overhead, throughput, and delay.

Biography:

Nei Kato is a full professor and the Director of Research Organization of Electrical Communication(ROEC), Tohoku University, Japan. He has been engaged in research on computer networking, wireless mobile communications, satellite communications, ad hoc & sensor & mesh networks, smart grid, IoT, Big Data, and pattern recognition. He has published more than 350 papers in prestigious peer-reviewed journals and conferences. He is the Editor-in-Chief of IEEE Network Magazine(2015.7-), the Associate Editor-in-Chief of IEEE Internet of Things Journal(2013-), and the Chair of IEEE Communications Society Sendai Chapter. He served as a Member-at-Large on the Board of Governors, IEEE Communications Society(2014-2016), a Vice Chair of Fellow Committee of IEEE Computer Society(2016),  a member of IEEE Computer Society Award Committee(2015-2016) and IEEE Communications Society Award Committee(2015-2017). He has also served as the Chair of Satellite and Space Communications Technical Committee(2010-2012) and Ad Hoc & Sensor Networks Technical Committee(2014-2015) of IEEE Communications Society. His awards include Minoru Ishida Foundation Research Encouragement Prize(2003), Distinguished Contributions to Satellite Communications Award from the IEEE Communications Society, Satellite and Space Communications Technical Committee(2005), the FUNAI information Science Award(2007), the TELCOM System Technology Award from Foundation for Electrical Communications Diffusion(2008), the IEICE Network System Research Award(2009), the IEICE Satellite Communications Research Award(2011), the KDDI Foundation Excellent Research Award(2012), IEICE Communications Society Distinguished Service Award(2012), IEICE Communications Society Best Paper Award(2012), Distinguished Contributions to Disaster-resilient Networks R&D Award from Ministry of Internal Affairs and Communications, Japan(2014), Outstanding Service and Leadership Recognition Award 2016 from IEEE Communications Society Ad Hoc & Sensor Networks Technical Committee, Radio Achievements Award from Ministry of Internal Affairs and Communications, Japan (2016) and Best Paper Awards from IEEE ICC/GLOBECOM/WCNC/VTC. Nei Kato is a Distinguished Lecturer of IEEE Communications Society and Vehicular Technology Society. He is a fellow of IEEE and IEICE.

HIT SEIE Summer Academic Activities IV: Invited Talk by Prof. Xuemin (Sherman) Shen

Harbin Institute of Technology Advanced Communications Technologies Forum 2017

Title:Connected Vehicles for Modern Transportation Systems

Date and Time: 8:30—10:00, Auguest 8, 2017

Location:ROOM 1013, BUILDING 2A, NO.2 YIKUANG STREET,

HARBIN, HEILONGJIANG, CHINA

Abstract:Modern society depends on faster, safer, and environment friendly transportation system. Vehicular communications network in terms of vehicle to vehicle, vehicle to infrastructure, vehicle to pedestrian, vehicle to cloud, and vehicle to sensor, provides a solution to modern transportation. In this talk, we first introduce the all connected vehicles. We then present the challenges and scientific research issues. of vehicular communications network. As examples, we show how to utilize mobility characteristics of vehicles to derive the achievable asymptotic throughput capacity in VANETs, and how to develop the charging strategies based on mobility of electric vehicles to improve the electricity utility, in order to approach load capacities of charging stations in VANET-enhanced smart grid. We conclude the talk by discuss the future autonomous driving.

Biography:

Xuemin (Sherman) Shen is a University Professor  and Associate Chair for Graduate Studies, Department of Electrical and Computer Engineering, University of Waterloo, Canada. Dr. Shen’s research focuses on wireless resource management, wireless network security, wireless body area networks, smart grid and vehicular ad hoc and sensor networks. He is the Editor-in-Chief of IEEE IoT Journal. He serves as the General Chair for Mobihoc’15, the Technical Program Committee Chair for IEEE GC’16, IEEE Infocom’14, IEEE VTC’10, the Symposia Chair for IEEE ICC’10, the Technical Program Committee Chair for IEEE Globecom’07, the Chair for IEEE Communications Society Technical Committee on Wireless Communications. Dr. Shen was an elected member of IEEE ComSoc BoG, the chair of IEEE ComSoc Distinguish Lecturer selection committee, and a member of IEEE ComSoc Fellow evaluation committee. Dr. Shen received the Excellent Graduate Supervision Award in 2006, and the Premier’s Research Excellence Award (PREA) in 2003 from the Province of Ontario, Canada. Dr. Shen is a registered Professional Engineer of Ontario, Canada, an IEEE Fellow, an Engineering Institute of Canada Fellow, a Canadian Academy of Engineering Fellow, a Royal Society of Canada Fellow, and a Distinguished Lecturer of IEEE Vehicular Technology Society and Communications Society.

HIT SEIE Summer Academic Activities III: Invited Talk by Dr. Mianxiong Dong

Title:Human-Like Driving: Empirical Decision-Making System for Autonomous Vehicles

Date and Time: 15:30—17:00, July 31, 2017

Location:ROOM 1013, BUILDING 2A, NO.2 YIKUANG STREET,

HARBIN, HEILONGJIANG, CHINA

Abstract:

Autonomous vehicle, as an emerging and rapidly growing field, has received extensive attention for its futuristic driving experiences. Although the rise of depth sensor technologies and machine learning methods have given a huge boost to self-driving research, existing autonomous driving vehicles do meet with several avoidable accidents during their road testing. The major cause is the misunderstanding between self-driving systems and human drivers. To solve this problem, we propose a human-like driving system in the paper to give autonomous vehicles the ability to make decisions like a human. In our method, a Convolutional Neural Network (CNN) model is used to detect, recognize and abstract the information in the input road scene, which is captured by the on-board sensors. And then a decision-making system calculates the specific commands to control the vehicles based on the abstractions. The most significant advantage in our work is that the proposed method can well adapt to real-life road conditions, in which a massive number of human drivers exist. In addition, we build our perception system only on the depth information, and avoid the unstable RGB data. Simulations demonstrate that our approach is robust and efficient, and outperforms the state-of-the-art in several related tasks.

 

Biography: Mianxiong Dong received B.S., M.S. and Ph.D. in Computer Science and Engineering from The University of Aizu, Japan. He is currently an Associate Professor in the Department of Information and Electronic Engineering at the Muroran Institute of Technology, Japan. Prior to joining Muroran-IT, he was a Researcher at the National Institute of Information and Communications Technology (NICT), Japan. He was a JSPS Research Fellow with School of Computer Science and Engineering, The University of Aizu, Japan and was a visiting scholar with BBCR group at University of Waterloo, Canada supported by JSPS Excellent Young Researcher Overseas Visit Program from April 2010 to August 2011. Dr. Dong was selected as a Foreigner Research Fellow (a total of 3 recipients all over Japan) by NEC C&C Foundation in 2011. His research interests include Wireless Networks, Cloud Computing, and Cyber-physical Systems. He has received best paper awards from IEEE HPCC 2008, IEEE ICESS 2008, ICA3PP 2014, GPC 2015, IEEE DASC 2015 and IEEE VTC 2016-Fall. Dr. Dong serves as an Editor for IEEE Communications Surveys and Tutorials, IEEE Network, IEEE Wireless Communications Letters, IEEE Cloud Computing, IEEE Access, and Cyber-Physical Systems (Taylor & Francis), as well as a leading guest editor for ACM Transactions on Multimedia Computing, Communications and Applications (TOMM), IEEE Transactions on Emerging Topics in Computing (TETC), IEEE Transactions on Computational Social Systems (TCSS), Peer-to-Peer Networking and Applications (Springer) and Sensors, as well as a guest editor for Concurrency and Computation: Practice and Experience (Wiley), IEEE Access, Peer-to-Peer Networking and Applications (Springer), IEICE Transactions on Information and Systems, and International Journal of Distributed Sensor Networks. He has been serving as the Vice Chair of IEEE Communications Society Asia/Pacific Region Meetings and Conference Committee, Program Chair of IEEE SmartCity 2015 and Symposium Chair of IEEE GLOBECOM 2016, 2017. Dr. Dong was a research scientist with A3 Foresight Program (2011-2016) funded by Japan Society for the Promotion of Sciences (JSPS), NSFC of China, and NRF of Korea. He is the recipient of IEEE TCSC Early Career Award 2016.

 

HIT SEIE Summer Academic Activities II: Invited Talk by Dr. Kaoru Ota

Title:Eyes in the Dark: Distributed Scene Understanding for Disaster Management

Date and Time: 13:30—15:30, July 31, 2017

Location:ROOM 1013, BUILDING 2A, NO.2 YIKUANG STREET,

HARBIN, HEILONGJIANG, CHINA

Abstract:

Robotic is a great substitute for human to explore the dangerous areas, and will also be a great help for disaster management. Although the rise of depth sensor technologies gives a huge boost to robotic vision research, traditional approaches cannot be applied to disaster-handling robots directly due to some limitations. In this paper, we focus on the 3D robotic perception, and propose a view-invariant Convolutional Neural Network (CNN) Model for scene understanding in disaster scenarios. The proposed system is highly distributed and parallel, which is of great help to improve the efficiency of network training. In our system, two individual CNNs are used to, respectively, propose objects from input data and classify their categories. We attempt to overcome the difficulties and restrictions caused by disasters using several specially-designed multi-task loss functions. The most significant advantage in our work is that the proposed method can learn a view-invariant feature with no requirement on RGB data, which is essential for harsh, disordered and changeable environments. Additionally, an effective optimization algorithm to accelerate the learning process is also included in our work. Simulations demonstrate that our approach is robust and efficient, and outperforms the state-of-the-art in several related tasks.

 

Biography: Kaoru Ota was born in Aizu-Wakamatsu, Japan. She received M.S. degree in Computer Science from Oklahoma State University, USA in 2008, B.S. and Ph.D. degrees in Computer Science and Engineering from The University of Aizu, Japan in 2006, 2012, respectively. She is currently an Assistant Professor with Department of Information and Electronic Engineering, Muroran Institute of Technology, Japan. From March 2010 to March 2011, she was a visiting scholar at University of Waterloo, Canada. Also she was a Japan Society of the Promotion of Science (JSPS) research fellow with Kato-Nishiyama Lab at Graduate School of Information Sciences at Tohoku University, Japan from April 2012 to April 2013. Her research interests include Wireless Networks, Cloud Computing, and Cyber-physical Systems. Dr. Ota has received best paper awards from ICA3PP 2014, GPC 2015, IEEE DASC 2015, IEEE VTC 2016-Fall and FCST 2017. She is an editor of IEEE Transactions on Vehicular Technology (TVT), IEEE Communications Letters, Peer-to-Peer Networking and Applications (Springer), Ad Hoc & Sensor Wireless Networks, International Journal of Embedded Systems (Inderscience) and Smart Technologies for Emergency Response & Disaster Management (IGI Global), as well as a guest editor of ACM Transactions on Multimedia Computing, Communications and Applications (leading), IEEE Communications Magazine, IEEE Network, etc. Also she was a guest editor of IEEE Wireless Communications (2015), IEICE Transactions on Information and Systems (2014), and Ad Hoc & Sensor Wireless Networks (Old City Publishing) (2014). She was a research scientist with A3 Foresight Program (2011-2016) funded by Japan Society for the Promotion of Sciences (JSPS), NSFC of China, and NRF of Korea.