Maxim Kontsevich is a permanent professor at the Institut des Hautes Études Scientifiques (IHÉS), holding the AXA Chair for Mathematics since 1995 and a visiting chair at Rutgers University (one month annually since 1997). Born in 1964 in Khimki, USSR, he earned his PhD from Bonn University in 1992. His career includes visiting positions at Harvard, the Institute for Advanced Study, and Berkeley, where he was a professor from 1993 to 1995. His research spans mathematical physics, algebraic geometry, and non-commutative geometry. Notable contributions include deformation quantization, mirror symmetry, and motivic integration. His work bridges algebraic structures with geometric and physical concepts, influencing areas like topological field theories, string theory, and integrable systems. Awardees of Fields Medal (1998), Crafoord Prize (2008), and Breakthrough Prize (2014), he also holds editorial roles at Compositio Mathematica and Publications Mathématiques IHÉS. His over 50 publications explore advanced topics such as quantum cohomology, Hodge theory, and categorical structures in geometry.
Xia Ben Hu is a Professor in the Department of Computer Science at Rice University's Brown School of Engineering. He leads research in automated and interpretable machine learning algorithms with applications across social informatics, health informatics, and information security. His work has resulted in widely adopted systems including AutoKeras, TODS, and RLCard. Education: PhD from Arizona State University (supervised by Dr. Huan Liu) Master and Bachelor degrees from Beihang University Prof. Hu's research focuses on developing automated and interpretable machine learning algorithms for large-scale, networked, dynamic and sparse data. His work spans automated machine learning (AutoML), deep learning, fairness in AI, time series analysis, and interpretable AI. He has made significant contributions to neural architecture search, collaborative filtering, anomaly detection, and reinforcement learning in imperfect information games. His publication record shows a clear progression from foundational work in network embedding and collaborative filtering (2017) to more recent work on LLM optimization, quantization, and extending context windows (2024). A consistent thread throughout his research is the focus on making complex machine learning systems more accessible, efficient, and interpretable. Selected Awards: ACM SIGKDD Rising Star Award (2021) NSF CAREER Award (2018) Teaching + Research Excellence Award, Rice University (2023) Multiple Best Paper Awards at top venues including ICML, CIKM, and AMIA Prof. Hu has successfully mentored numerous graduate students, with recent graduates securing tenure-track positions at major universities. His research is generously supported by federal agencies including DARPA (XAI, D3M, NGS2), NSF (CAREER, III, SaTC), NIH, and industrial sponsors such as Adobe, Apple, Google, LinkedIn, and JP Morgan. He has served as General Co-Chair for WSDM 2020 and ICHI 2023, and Program Chair for AIHC 2024. He leads the DATA Lab at Rice University, which develops open-source systems for automated machine learning and reinforcement learning. The lab's AutoKeras system has over 8,000 GitHub stars and 1,000 forks, and their work has been integrated into TensorFlow, Apple production systems, and Bing production systems.
Andrea Goldsmith is the Dean of the School of Engineering and Applied Science and the Arthur LeGrand Doty Professor of Electrical and Computer Engineering at Princeton University. Previously, she held the Stephen Harris Professorship at Stanford University and remains Harris Professor Emerita there. Her research focuses on information theory, communication theory, signal processing, and their applications to wireless communications, interconnected systems, and neuroscience. She founded Plume WiFi and Quantenna, Inc., and serves on the boards of Medtronic and Crown Castle Inc. Education: B.S., M.S., and Ph.D. in Electrical Engineering, University of California, Berkeley (1986–1994) Research Interests: Her work bridges theoretical foundations with practical applications in wireless systems, including MIMO communications, cognitive radio, and the integration of machine learning in communication protocols. She also explores the intersection of wireless technology with biomedical systems and neuroscience, emphasizing innovations like smart buildings and in-body networks. Key Contributions: Authored seminal textbooks, including Wireless Communications and MIMO Wireless Communications . Inventor on 29 patents, with significant industry impact through startups. Recipient of prestigious awards such as the IEEE Sumner Award, ACM Athena Lecturer Award, and Marconi Prize. Labs & Leadership: Leads the Wireless Systems Lab at Princeton, advancing cutting-edge wireless technologies. Chair of the IEEE Board of Directors Committee on Diversity, Inclusion, and Ethics.
Anshumali Shrivastava is an Associate Professor of Computer Science, Electrical and Computer Engineering, and Statistics at Rice University, affiliated with the George R. Brown School of Engineering. His research focuses on large-scale machine learning, randomized algorithms for big data, and graph mining. He holds a PhD from Cornell University (2015) and an MSc from the Indian Institute of Technology Kharagpur (2008). His research interests span scalable deep learning, efficient neural network inference, and probabilistic algorithms. He has pioneered techniques in compressed learning, hashing-based search, and distributed optimization for handling massive datasets. Notable contributions include methods for accelerating LLM inference, memory-efficient quantization, and graph processing algorithms. Teaching: Probabilistic Algorithms, Large-Scale ML, and Machine Learning Seminars Awards: Charles W. Duncan Jr. Achievement Award (2023), Young Faculty Research Award (2021), NSF CAREER Award (2017), and multiple best paper awards His work bridges algorithm design with practical applications in recommendation systems, genomics, and edge computing. Current efforts focus on sustainable AI, hardware-aware compression, and efficient training/inference pipelines for large models.
Svetlana Lazebnik is a Full Professor and Willett Faculty Scholar in the Department of Computer Science at the University of Illinois at Urbana-Champaign (UIUC), part of the Grainger College of Engineering. She holds a Ph.D. from UIUC (2006) and previously served as an Assistant Professor at the University of North Carolina at Chapel Hill (2007–2011). Her research focuses on computer vision, including generative models for virtual try-on, image stylization, scene understanding, and joint modeling of images and language. She has advised numerous Ph.D. students and postdocs, many of whom now hold prominent academic and industry roles. Education: Ph.D. in Computer Science, UIUC (2006); supervised by Jean Ponce. Research Interests: Her work spans generative adversarial networks (GANs), diffusion models, virtual try-on systems (e.g., Dressing-in-Order, Street Try-On), exemplar-based stylization, and large-scale photo analysis. She has pioneered spatial pyramid matching and contributed to binary code learning for image retrieval. Key Awards: NSF CAREER Award (2008), Microsoft Research Faculty Fellow (2009), Sloan Research Fellow (2013), IEEE Fellow (2021), and the Longuet-Higgins Prize (2016) for her CVPR 2006 paper. Teaching: Recent courses include CS 444 (Deep Learning for Computer Vision), CS 543 (Computer Vision), and a Ph.D. Job Search Seminar. She has also taught at UNC Chapel Hill. Grants & Funding: Supported by NSF, Amazon, AWS, Microsoft, Sloan Foundation, Google, ARO, and Adobe. Notable grants include CCF 2348624 and IIS 1718221. Labs/Groups: Leader in the Illinois CS Vision Group, contributing to collaborative projects on embodied AI, multi-agent systems, and visual-semantic reasoning.
Lizi Liao is an Assistant Professor at the School of Computing and Information Systems , Singapore Management University (SMU) , specializing in Artificial Intelligence and Conversational AI . Her research bridges Machine Learning , Natural Language Processing , and Multimodal Systems , focusing on proactive dialogue systems, multimodal conversational search, and task-oriented interactions. Education : PhD in Computer Science (2019) from the National University of Singapore (NUS) , advised by Professor Tat-Seng Chua . Research Interests center on principles of human conversational understanding and machine implementation, particularly in proactive conversational agents , multimodal dialogue systems , and target-driven conversation planning . Key applications include emotional support systems , intelligent shopping assistants , and learning companions . Recent Publications (2024-2025) highlight her work on LLM-based proactive dialogue , multimodal emotion recognition , and dynamic graph modeling , often integrating NLP , Multimedia , and Knowledge Graphs . Collaborative projects with her CoAgent Lab team emphasize human-AI interaction and ethical response generation . Scientific Awards : Google South Asia & Southeast Asia Research Award 2023 Lee Kong Chian Fellow Teaching includes Visual Analytics for Business Intelligence (undergraduate) and Text Analytics and Application (graduate). She also serves as Associate Editor for TOIS and TOMM , and organizes tutorials at ACL , SIGIR , and WSDM .
Ali Ghodsi is a Professor at the University of Waterloo and Director of the Data Science Lab, with affiliations at the Vector Institute. His research spans machine learning, deep learning, and artificial intelligence, with applications in natural language processing, bioinformatics, and computer vision. His group develops theoretical frameworks and algorithms for analyzing large-scale datasets, focusing on neural network architectures, knowledge distillation, and model efficiency. Current projects include deep learning for identity control, computational antibody design, and generative AI/large language models. Ghodsi has authored influential tutorials on diffusion models, graph neural networks, and large language models. Notable research contributions include computational methods for de novo peptide sequencing from mass spectrometry data, green simulation-assisted reinforcement learning, and efficient natural language processing models. His lab maintains collaborations with industry partners including Google, Amazon, and Roche.
Dr. Hyung Jin Chang is an Associate Professor at the School of Computer Science, University of Birmingham, and a Turing Fellow at the Alan Turing Institute. He holds a Ph.D. and B.S. from Seoul National University. His research focuses on human-centered visual learning, particularly in human-robot interaction, with expertise in computer vision, machine learning, and deep learning. He has been involved in organizing conferences like ECCV and ICCV workshops (e.g., VOTS Challenge, HANDS Workshop) and serves on program committees including AAAI and CVPR. His work spans areas like gaze estimation, domain adaptation, 3D pose estimation, and robotic perception for assistive technologies. Key achievements include receiving the Royal Society Research Grant (2019–2020) and Wellcome Trust funding. Notable contributions include frameworks for unsupervised domain adaptation, gaze estimation models (e.g., RT-Gene), and collaborative learning methods for hand-object reconstruction. He has led projects in medical robotics, personalized dressing assistance, and safety-critical systems like driver attention prediction. His 15 most recent articles (2024–2025) emphasize advancements in diffusion models, domain adaptation, 3D reconstruction, gaze-controllable systems, and generative AI for motion and interaction modeling. These reflect a trend toward integrating multimodal data (vision + language) and bridging theoretical foundations with applied robotics. Awards: Royal Society Grant, Wellcome Trust, Turing Fellowship Grants: Active in securing funding for robotics, vision, and healthcare applications He leads the Personal Robotics Lab and collaborates on projects like the VOTS Challenge for visual object tracking. His research bridges academia and real-world applications in healthcare robotics and human-technology interaction.
Esa Ollila serves as Associate Professor in the Department of Signal Processing and Acoustics at Aalto University, Finland, and holds an adjunct professorship in Statistics at the University of Oulu. His academic appointments include Academy of Finland Research Fellow (2010-2015) and prior senior research/lecturing roles at both institutions. His educational background features: M.Sc. in Mathematics, University of Oulu (1998) Ph.D. in Statistics (with honors), University of Jyväskylä (2002) D.Sc.(Tech) in Signal Processing (with honors), Aalto University (2010) Professor Ollila's research centers on statistical signal processing and robust statistical methodologies , with significant contributions to array processing, high-dimensional data analysis, and covariance matrix estimation. His work bridges theoretical statistics with practical applications in radar systems, wireless communications, and big data analytics, emphasizing robustness against outliers and computational efficiency in modern data-intensive environments. Current focus areas include compressed sensing, sparse approximation, and blind source separation techniques. Analysis of his 15 most recent publications (2024-2025) reveals three dominant trends: (1) robust covariance learning for massive random access systems, (2) integrated sensing and communications (ISAC) for 6G networks using advanced beamforming, and (3) geometric approaches to elliptical distributions in statistical inference. His work increasingly incorporates deep learning (GANs, graph neural networks) while maintaining strong foundations in classical signal processing theory. Key recognitions include: Academy of Finland Postdoctoral Fellowship (2004-2007) Academy of Finland Research Fellowship (2010-2015) His research has been supported through prestigious Academy of Finland grants totaling over a decade of continuous funding. Professor Ollila currently leads an active research group at Aalto University, supervising doctoral candidates and collaborating internationally with institutions including Princeton University (where he served as Visiting Post-doctoral Research Associate during 2010-2011). He maintains strong ties with the University of Oulu through his adjunct professorship and has contributed to EURASIP's Special Area Team on Theoretical and Methodological Trends in Signal Processing. The Esa Ollila Research Group focuses on cutting-edge challenges in statistical signal processing, with current projects spanning robust DOA estimation under non-Gaussian noise, covariance matrix learning for massive MIMO systems, and machine learning-enhanced radar-communication integration. The group actively develops open-source tools like the fitHeavyTail R package for heavy-tailed distribution modeling and maintains collaborations with industry partners in wireless communications.
HaoYu Wang is an Assistant Professor of Computer Science at SUNY Albany. His research focuses on parameter-efficient and data-efficient deep learning, particularly in natural language processing and machine learning, aiming to democratize AI access. He holds a Ph.D. from Purdue University's School of Electrical and Computer Engineering, a B.Eng. from the University of Electronic Science and Technology of China, and an MS from SUNY Buffalo. His work includes innovations like RoseLoRA (sparse low-rank adaptation for knowledge editing), LightLT (lightweight quantization for long-tail data), and FedKC (federated knowledge composition for multilingual NLU). He has received awards such as the Future Leaders in Data Science (2024) and Bilsland Dissertation Fellowship. Key research areas include parameter efficiency, cross-lingual NLU, and model fairness. Recent publications span topics like federated learning optimization, robust retrieval-augmented generation, and mitigating token overfitting in LLMs. He advises students on Ph.D. and intern roles, emphasizing CVs and research interests in applications.
Professor David Taubman is a distinguished academic serving as Professor and Deputy Head of School (Research) at the School of Electrical Engineering and Telecommunications (EE&T) at the University of New South Wales (UNSW) in Sydney, Australia. He is also co-director of Kakadu Software Pty. Ltd. and its affiliates Kakadu R&D and Kakadu GPU. With a career spanning over three decades, Professor Taubman has made significant contributions to the field of image and video compression, most notably as the author of the EBCOT coding algorithm adopted in the JPEG2000 international standard. Professor Taubman earned his B.Sc. in Mathematics and Computer Science (1986) and B.E. (Medal) in Electrical Engineering (1988) from the University of Sydney, followed by an M.Sc. (1992) and Ph.D. (1994) in Electrical Engineering from the University of California at Berkeley. His professional journey includes engineering work at the Electricity Commission of N.S.W. (1988-1990), research positions at Hewlett-Packard Laboratories in Palo Alto (1994-1998), and an academic career at UNSW where he progressed from Senior Lecturer (1998-2003) to Associate Professor (2004-2009) and finally to Professor (2009-present). He has held various leadership roles including Head of the EE&T Telecommunications Research Group (2003-2014), Head of the EE&T Signal Processing Research Group (2014-present), Director of Research for the School of EE&T (2011-2016), and Deputy Head of School (Research) since 2017. Professor Taubman's research interests center on image and video compression, with particular expertise in JPEG2000 standards and implementations. His work spans signal processing, wavelet transforms, scalable video coding, motion modeling, and multimedia systems. He has pioneered numerous compression algorithms and frameworks, including the EBCOT coding algorithm that became central to the JPEG2000 standard. His recent research focuses on efficient motion modeling with cuboidal partitioning, learned lifting-based transform structures, and high-throughput implementations of JPEG2000 for video applications. His work bridges theoretical foundations with practical implementations, as evidenced by the commercially successful Kakadu Software tools that have garnered around 500 commercial licensees. Analysis of Professor Taubman's recent publications reveals a consistent focus on advancing compression technologies with particular emphasis on scalability, efficiency, and adaptability. His work spans traditional image compression (JPEG2000 extensions), video coding (cuboid-based partitioning for UHD/360-degree video), and emerging applications (nanopore sequencing data compression). A notable trend is the integration of machine learning techniques with traditional compression frameworks, as seen in his work on learned lifting-based transform structures. His research maintains strong connections to real-world applications across diverse domains including medical imaging, astronomical data processing, and genomic sequencing. IEEE Fellow Engineers Australia Fellow (by invitation) Professor Taubman has served as Associate Editor for the IEEE Transactions on Image Processing for two four-year appointments (2003-2005 and 2010-2013). He has been actively involved in numerous research grants focused on image and video compression technologies, particularly those related to the JPEG2000 standard and its extensions. His work has received significant industry support, reflected in his consultancy with various U.S., Japanese, and Australian corporations. He has also contributed to international standards development as a member of Standards Australia Technical Committee MS-065 (mirroring ISO TC42 on Digital Photography) and as a constitutional member of Standards Australia Technical Committee IT-029 (Coded Representation of Picture, Audio and Multimedia/Hypermedia Information). Professor Taubman co-directs Kakadu Software Pty. Ltd. and its research affiliates Kakadu R&D and Kakadu GPU, which have developed the commercially successful Kakadu Software tools for JPEG2000. His research group at UNSW focuses on advanced image and video compression techniques, with particular expertise in wavelet-based methods, scalable coding, and motion modeling. The group maintains strong industry connections and has contributed significantly to the development and standardization of image compression technologies worldwide.
Qiongxiu Li is a Tenure-Track Assistant Professor in the Cyber Security group at Aalborg University's Copenhagen campus, part of the Technical Faculty of IT and Design. Her research focuses on cybersecurity, distributed optimization, privacy/security, and federated learning. She has authored/co-authored 38 papers in top-tier venues including IEEE Transactions on Information Forensics and Security, ICLR, and EUSIPCO. Education: PhD in Privacy and Security from Aalborg University (2018-2021). Notable achievements include winning the EUSIPCO 2020 3MT Contest and co-delivering a tutorial on privacy-preserving distributed optimization at EUSIPCO 2024. She actively reviews for conferences like NeurIPS, ICLR, and journals such as TPAMI and TIFS. Research Themes: Privacy-preserving distributed algorithms, federated learning security, differential privacy, and adversarial machine learning. Recent Trends: Focus on securing AI systems (e.g., LLM vulnerabilities, federated clustering privacy), quantization for privacy, and theoretical bounds in decentralized learning. Awards: 2020 EUSIPCO 3MT Winner (outstanding finalist in EURASIP's annual doctoral research competition). Grants/Projects: Co-PI of the AI:SECURITY project (2025-2029) addressing AI security threats like phishing and malicious actors. Labs/Teams: Leads the Cyber Security group at Aalborg's Copenhagen campus, focusing on theoretical and applied research in secure distributed systems.
Arya Mazumdar is a tenured Professor of Data Science and Computer Science at the Halıcıoğlu Data Science Institute (HDSI) , part of the School of Computing, Information and Data Sciences at the University of California San Diego (UCSD). He also holds affiliations with the Computer Science and Engineering and Electrical and Computer Engineering departments at UCSD. Previously, he was an Assistant and Associate Professor at the University of Massachusetts Amherst (2015–2021), and held a postdoctoral position at MIT (2011–2012). He is an IEEE Distinguished Lecturer (2023–2024) and recipient of the NSF CAREER Award (2015–2020). Education : PhD in 2011 from the University of Maryland, College Park (advisor: Alexander Barg) Postdoctoral scholar at MIT (2011–2012, advisor: Greg Wornell) Internships at IBM Almaden (2010) and HP Labs (2008) Research Interests : Focus on algorithmic and statistical aspects of machine learning, error-correcting codes, optimization, signal processing, and distributed systems. Key areas include clustering algorithms, compressed sensing, federated learning, and information-theoretic foundations of data science. He has contributed to theoretical guarantees for learning mixtures, distributed optimization, and robust coding schemes. Recent Articles Trends : Recent work emphasizes distributed optimization (e.g., vqSGD, Byzantine-resilient algorithms), sparse recovery in high-dimensional models, and theoretical foundations of learning (e.g., support recovery, parameter estimation). His research bridges information theory and machine learning, with applications in storage systems and large-scale data processing. Awards & Roles : 2020 EURASIP JASP Best Paper Award Co-PI and co-leader of the NSF AI Institute for Learning-Enabled Optimization at Scale Editorships: IEEE Transactions on Information Theory and Foundations and Trends in Communications Grants & Labs : Leads the EnCORE Institute as UCSD Site Lead, focusing on theoretical perspectives of large language models and computational-statistical gaps. Active in organizing workshops on topics like LLMs, clustering, and distributed optimization. His research is funded by NSF and industry collaborations. Teaching : Courses include Algorithms for Data Science , Probability and Statistics for Data Science , and Coding Theory , emphasizing foundational theory and scalable methods.
Sandra Keiper is a Lecturer at the Institute of Mathematics within Faculty II - Mathematics and Natural Sciences at Technical University of Berlin. She has held academic positions since at least 2011, including roles as Tutor, Assistant, and Lecturer, with teaching responsibilities in Analysis, Linear Algebra, and Partial Differential Equations for both mathematicians and engineers. Research interests include: Compressed Sensing and Sparse Signal Recovery Numerical Linear Algebra with applications to high-dimensional data Wavelet and curvelet transforms for geometric multiscale analysis Approximation theory for finite-valued and cartoon-like functions Deep learning and graph approximation techniques Professional activities : Active in teaching since 2011 (Analysis I-III, Functional Analysis, Integral Transforms) Supervising theses since 2015 on topics like Compressed Sensing and Deep Learning Invited lectures at Caltech, ETH Zurich, and Alan Turing Institute Research stays at Hausdorff Institute, ETH Zurich, and Duke University
Luciano Lavagno is a Full Professor at the Department of Electronics and Telecommunications, Polytechnic University of Turin, with over two decades of academic and research contributions. His work bridges hardware acceleration, low-power electronics, and intelligent system design. Research Focus: Hardware-accelerated machine learning, high-level synthesis (HLS) for FPGA/ASIC, heterogeneous CPU/GPU/FPGA platforms Key Projects: SPACE (predictable acceleration), REBECCA (secure AI acceleration), HPC-National Center (quantum computing), and oral history preservation via "Ti racconto una storia" initiative His recent publications analyze CNN inference optimization, subgraph isomorphism, and superword-level parallelism exploitation. Lavagno supervises multiple PhD students working on FPGA acceleration, neural network hardware, and embedded systems. As Principal Investigator for national and EU-funded projects (PRIN, JTI-ECSEL, PNRR), he drives advancements in sustainable computing infrastructure. His patented technologies include MIx&Latch timing methodology, capacitive sensing innovations, and 5G acceleration frameworks.
.jpg)
