KEYNOTE SERIES

                                    Zhongqi Miao

                             (Applied Research Scientist, Microsoft)

Bio: Zhongqi Miao is an Applied Research Scientist at Microsoft’s AI for Good Research Lab, where he focuses on incorporating artificial intelligence into environmental science and ecology domains. He obtained his Ph.D. from UC Berkeley under the mentorship of Prof. Wayne Getz and Prof. Stella Yu, specializing in computer vision and deep learning applications, particularly in camera trap recognition. Currently, Zhongqi broadens his expertise in wildlife recognition from diverse sources such as ground-based imagery, bio-acoustics, and aerial imagery from planes, drones, and satellites. Motivated by real-world challenges, he is also enthusiastic about addressing complex issues like long-tail distribution, multi-domain applications, human-in-the-loop interactions, and enhancing the deployment efficiency of large foundation models.

Title of the talk: Challenges and solutions of deep learning in wildlife conservation

Abstract: Deep learning has garnered significant interest from the ecological community due to its ability to extract and generalize patterns from complex datasets, such as images, audio, and motion signals. However, despite its potential, deep learning may exhibit limitations when applied to real-world ecological datasets.

In this presentation, I will outline four characteristics of realistic datasets: 1) long-tailed; 2) open-ended; 3) multi-domain; and 4) imbalance between labeled vs. unlabeled data. I will then focus on one of our recent papers, “Iterative Human and Automated Identification of Wildlife Images,” which demonstrates how AI recognition systems can be made deployable with efficient human-in-the-loop and continuous domain adaptation, considering the advancements of state-of-the-art Large Foundation Models.

Ling Liu

(Professor, Georgia Institute of Technology, USA)

Bio: Ling Liu Ling Liu is a Professor in the School of Computer Science at Georgia Institute of Technology. She directs the research programs in the Distributed Data Intensive Systems Lab (DiSL), examining various aspects of large scale big data-powered artificial intelligence (AI) systems, and machine learning (ML) algorithms and analytics, including performance, availability, privacy, security and trust. Prof. Liu is an elected IEEE Fellow, a recipient of IEEE Computer Society Technical Achievement Award (2012), and a recipient of the best paper award from numerous top venues, including IEEE ICDCS, WWW, ACM/IEEE CCGrid, IEEE Cloud, IEEE ICWS. Prof. Liu served on editorial board of over a dozen international journals, including the editor in chief of IEEE Transactions on Service Computing (2013-2016), and the editor in chief of ACM Transactions on Internet Computing (since 2019). Prof. Liu is a frequent keynote speaker in top-tier venues in Big Data, AI and ML systems and applications, Cloud Computing, Services Computing, Privacy, Security and Trust. Her current research is primarily supported by USA National Science Foundation under CISE programs, IBM and CISCO.

Title of the talk: Security and Privacy in Federated Learning

Abstract:  Federated learning (FL) is an emerging distributed collaborative learning paradigm by decoupling the learning task from the centralized server to a decentralized population of edge clients. One of the attractive features of federated learning is its default client privacy, allowing clients to keep their sensitive training data locally and only share local model updates with the federated server. However, recent studies have revealed that such default client privacy is insufficient for protecting the privacy of client training data from both gradient leakage attacks and data poisoning attacks. This keynote will describe gradient leakage attacks and data poisoning attacks and provide insights for designing effective privacy and security strategies for combating privacy leakage attacks and data poisoning attacks.

Pietro Liò

(Professor, University of Cambridge, UK)

Bio: Dr. Liò received the PhD degree in complex systems and non linear dynamics from the
School of Informatics, dept of Engineering, University of Firenze, Italy and the PhD degree in
theoretical genetics from the University of Pavia, Italy. He is currently a professor of computational biology with the Department of Computer Science and Technology, University of Cambridge and a member of the Artificial Intelligence Group.
He is also a member of the Cambridge Centre for AI in medicine, ELLIS (European Laboratory for Learning and Intelligent Systems), Academia Europaea, Asia Pacific Artificial Intelligence Association, . His research interests include graph representation learning, AI and Medicine, Systems Biology.

Title of the talk: AI and medicine: graph and hypergraph representation learning

Alfred O. Hero

(Professor, University of Michigan, USA)

Bio: Alfred O. Hero III is the John H. Holland Distinguished University Professor of Electrical Engineering and Computer Science and the R. Jamison and Betty Williams Professor of Engineering at the University of Michigan, Ann Arbor. His primary appointment is in the Department of Electrical Engineering and Computer Science and he also has appointments, by courtesy, in the Department of Biomedical Engineering and the Department of Statistics. He is currently on leave from the University of Michigan as a Program Director in the CISE Directorate at the National Science Foundation. He received the B.S. (summa cum laude) from Boston University (1980) and the Ph.D from Princeton University (1984), both in Electrical Engineering. He is a Life Fellow of the Institute of Electrical and Electronic Engineers (IEEE) and a Fellow of the Society for Industrial and Applied Mathematics (SIAM). He helped launch the SIAM Journal on Mathematics of Data Science and served as Section Editor until 2022. He has served as President of the IEEE Signal Processing Society and as a member of the IEEE Board of Directors. Alfred Hero has received numerous awards for his research and service to the profession including several best paper awards, the 2013 IEEE Signal Processing Society Technical Achievement Award, the 2015 Society Award from the IEEE Signal Processing Society and the 2020 Fourier Award from the IEEE. He received the 2017 Stephen S. Attwood Excellence in Engineering Award and the 2018 H. Scott Fogler Award for Professional Leadership and Service from the University of Michigan. Alfred Hero’s recent research interests are in modeling high dimensional spatio-temporal data, multimodal data integration, statistical signal processing, and machine learning.Of particular interest are predictive mathematical models for the biological and physical sciences, social networks, network security and forensics, and personalized health and disease.

Title of the talk: Wearable device networks for predicting health

Abstract: We are entering an era where diverse types of data can be collected continuously from our bodies using wearable devices. With the help of predictive models and algorithms this data can provide useful biomarkers for physiological and mental health. Some of these biomarkers can be used to predict susceptibility to disease and resilience to infection. In the future whole body networks of devices will be used to aggregate and analyze this data in real time, improving personal health awareness and well being. This talk will provide an overview of some opportunities and hurdles that will need to be overcome to make this a reality.

NGUYEN THANH TUNG

(CEO of Khai Minh Technology Group (KMTG))

Bio: Dr. Tung is the Rector/Vice Chairman of the Industrial University of Vinh City.  Previously he was the Rector at Vabis International College. He is also the CEO/Co-Founder of Khai Minh Technology Group and Tuệ Đức Green School System (20+ campus). He is having more than 29 years of experience in Multinational Companies & Education organisations. He has a Strong personality and Initiative, leadership capability and proactive leadership/ attitude.

Title of the talk: Customize AI model in business: Case study in KMTG Vietnam (Khai Minh Technology Group): KMTG’s Brain

Abstract: Customizing AI models for business is becoming increasingly important as more organizations seek to leverage the power of AI to gain a competitive advantage. While pre-built AI models can be effective in some cases, they are not always suitable for every business’s specific needs. By customizing AI models, businesses can tailor them to their specific requirements, improving their accuracy, speed, and relevance. It involves several steps to customizing AI models, including selecting the appropriate AI algorithms, collecting and preprocessing data, and fine-tuning the model to optimize performance.
This can be a complex and challenging process, requiring specialized skills and expertise. However, with the right approach and tools, businesses can successfully customize AI models to achieve their goals.
KMTG (Khai Minh Technology Group) on processing build the KMTG’s Brain by customizing AI models to improve efficiency, increase accuracy, reduce error rates. It can also help KMTG to identify new opportunities, reduce costs, and improve customer satisfaction. Despite the potential benefits, customizing AI models is not without its challenges. These challenges include data quality issues, the need for specialized skills and resources, and the need to continuously monitor and update the model to ensure it remains relevant and effective.
Overall, customizing AI models is a critical component of a successful AI strategy for businesses. By leveraging the power of AI and customizing models to their specific needs, businesses can gain a competitive edge and achieve their goals more effectively and efficiently.

SATYANDRA K. GUPTA

(Professor, University of Southern California, USA)

Bio: Dr. Satyandra K. Gupta holds Smith International Professorship in the Viterbi School of Engineering at the University of Southern California and serves as the Director of the Center for Advanced Manufacturing. He is also Co-Founder and Chief Scientist at GrayMatter Robotics. His research interests are physics-informed artificial intelligence, computational foundations for decision-making, and human-centered automation. He works on applications related to Manufacturing Automation and Robotics. He has published more than four hundred technical articles in journals, conference proceedings, and edited books. He is a fellow of the American Society of Mechanical Engineers (ASME), Institute of Electrical and Electronics Engineers (IEEE), Solid Modeling Association (SMA), and Society of Manufacturing Engineers (SME). He is a former editor-in-chief of the ASME Journal of Computing and Information Science in Engineering. He has received numerous honors and awards for his scholarly contributions. Representative examples include a Young Investigator Award from the Office of Naval Research in 2000, Robert W. Galvin Outstanding Young Manufacturing Engineer Award from the Society of Manufacturing Engineers in 2001, a CAREER Award from the National Science Foundation in 2001, a Presidential Early Career Award for Scientists and Engineers in 2001, Invention of the Year Award at the University of Maryland in 2007, Kos Ishii-Toshiba Award from ASME in 2011, Excellence in Research Award from ASME Computers and Information in Engineering Division in 2013, Distinguished Alumnus Award from Indian Institute of Technology, Roorkee in 2014, ASME Design Automation Award in 2021, and Distinguished Alumni Award from Indian Institute of Technology, Delhi in 2022. He was named “The 20 most influential professors in smart manufacturing” by Smart Manufacturing Magazine in June 2020. He has also received ten best paper awards at international conferences. He serves as a member of the Technical Advisory Committee for Advanced Robotics for Manufacturing (ARM) Institute and a member of the National Materials and Manufacturing Board.

Title of the talk: Physics – Informed AI for Enabling Robots to Learn Autonomous Tool Manipulation

Abstract: Humans’ ability to construct and use complex tools makes them different from animals. Many manufacturing applications such as sanding or composite layup require ergonomically challenging tool motions. Robots have successfully reduced the need for humans to perform tedious tasks in mass production applications. Robots are programmed by humans to execute pre-determined tool motions in mass production applications and human programming effort is amortized over a large number of parts. However, high-mix applications require the part changeover to be accomplished within a few minutes, therefore, we cannot rely on humans to program robots when a new part arrives. Unfortunately, at present, the use of robots in high-mix manufacturing applications is very limited, requiring humans to perform ergonomically challenging and physically demanding tasks. Using robots in these applications requires robots to autonomously manipulate tools based on high-level task descriptions and deliver human competitive task performance. This is a challenging problem and addressing it requires leveraging the latest advances in AI. This seminar will present an overview of physics-informed AI technologies that enable robots to learn safe and efficient autonomous tool manipulation. These new technologies serve as the foundation for realizing smart robotic cells for assembly, composite layup, additive manufacturing, inspection, and sanding applications. AI-based planning enables the automated generation of efficient robot trajectories for performing complex tool motions to meet process-specific requirements. Imitation learning enables robots to learn from human experts. The use of synthetic images generated from physics-informed simulations enables the use of deep learning in defect detection. Self-supervised active learning enables the robotic cell to autonomously and safely conduct experiments to learn the process parameters in the most efficient manner. Smart robotic cells increase human productivity and reduce the need for humans to perform ergonomically challenging tasks.

Victor Pankratius

(Vice President, AI & Data, Bosch Cross-Domain Computing Solutions)

Bio: Victor serves as a Vice President, AI & Data, in Bosch’s Cross-Domain Computing Solutions area. He is an experienced leader with MIT and NASA research background in artificial intelligence, data science, software engineering, and parallel computing. Prior to that he was the director and Head of Global Software Engineering at Bosch Sensortec, where he led a department breaking new ground in sensor software and Edge-AI in mobile devices, wearables, automotive, AR/VR, IoT, gaming, industrial, and environmental applications. Prior to Bosch, he led a data science group at MIT advancing computer-aided discovery & AI and served as a principal investigator in NASA’s prestigious Advanced Information Systems Technology program. Victor earned a Habilitation degree in Computer Science from KIT and a doctorate with distinction from the University of Karlsruhe’s business school. For more details, please visit: http://www.victorpankratius.com

Title of the talk: Trends in Sensing Applications and AI at the Edge

Abstract: New generations of sensors are increasingly equipped with microcontrollers and computing capabilities that enable local machine learning in millimeter-sized packages. This keynote presents use cases where sensing applications have become a major driver for Tiny AI. Applications are shown for intelligent Micro-Electro-Mechanical Systems (MEMS) in motion learning, sports analytics, and gas and environmental sensing. Looking at the software stack, this keynote also addresses the importance of formalizing and including domain knowledge into AI for optimizations, such as shrinking memory footprints, making trade-offs in signal processing, and algorithmic choice. Learning from individual success stories, our insights help sketch a bigger picture for AI-IoT ecosystems and platforms.

Rida Qadri

(Research Scientist, Google)

BIO: Rida Qadri is a Research Scientist in Google’s Responsible AI and Human-Centered Technology team. She leverages her interdisciplinary expertise and cross-cultural research experience to study the limitations of generative AI in non-western settings. Through this work, she seeks to build AI pipelines that are inclusive of global cultures and respect the situated expertise and knowledge of global communities. Her past research has examined  mobility platform and gig work algorithms in Jakarta, looking at the failures and frictions of these technologies in a non-western context. She completed her PhD in Computational Urban Science.