Bradley J. Nelson Switzerland, Eidgenössische Technische Hochschule Zürich
Member of the Serbian Academy of Sciences and Arts
Foreign Member of the Chinese Academy of Sciences

Member of the Swiss Academy of Engineering Sciences

Professor of Robotics and Intelligent Systems, ETH Zurich

Keynote Speech Title：Clinically Ready Microrobots
Abstract：Micro and nano robotics have advanced rapidly over the past two decades, driven by breakthroughs in functional materials, processing, and microfabrication, as well as improved modeling, sensing, and control. As a result, microrobots that are clinically relevant, biocompatible, and increasingly biodegradable, are now within reach, opening new possibilities for minimally invasive diagnostics and targeted therapy in anatomically confined and delicate environments.

Our group develops magnetic microrobots based on biocompatible composite materials and actuated by externally generated magnetic fields and gradients. Magnetic actuation offers several key advantages for clinical translation, including wireless power delivery, deep tissue penetration, and the ability to generate controlled forces and torques without onboard batteries or tethers. However, realizing robust and safe operation in vivo requires more than microrobot design alone. A central challenge is the development of magnetic navigation systems capable of generating accurate, programmable fields over human-scale workspaces while meeting clinical constraints such as patient access, imaging compatibility, and safety limits.

In this talk, I will present recent progress toward a clinically ready, human-scale microrobot platform, spanning microrobot fabrication, magnetic field generation and control, and closed-loop navigation strategies. I will also highlight emerging applications enabled by this platform, including concepts for telestroke medicine, where rapid, remote intervention could expand access to time-critical care. Together, these advances bring the long-standing “Fantastic Voyage” vision closer to practical medical reality.
Biography：Bradley J. Nelson has been Professor of Robotics and Intelligent Systems at ETH Zurich, Switzerland, since 2002. His research focuses on microrobotics, nanorobotics, and medical robotics. He has established an internationally renowned research group dedicated to translating fundamental scientific discoveries into practical robotic systems. His pioneering work has significantly advanced microrobotics, moving the field from conceptual demonstrations toward clinically relevant applications. His laboratory has also trained numerous researchers who now hold leading positions in academia, industry, and medicine.

Prof. Nelson is a Foreign Member of the Chinese Academy of Sciences, a member of the Swiss Academy of Engineering Sciences (SATW), and a Fellow of IEEE and ASME. His outstanding contributions to robotics, nanotechnology, and biomedicine have been recognized with numerous prestigious honors, including the IEEE Robotics and Automation Society Pioneer Award (2019), the Grand Hamdan International Award for AI in Healthcare (2020), and selection to the Scientific American 50 (2005).

Beyond his research achievements, Prof. Nelson has held several major academic leadership positions at ETH Zurich, including serving twice as Head of the Department of Mechanical and Process Engineering and as Chair of the ETH Electron Microscopy Center. He has also served as a member of the Research Council of the Swiss National Science Foundation.

Prof. Nelson is a founding editorial board member of Science Robotics and previously served as the journal's Chief Scientific Advisor.

A. Murat Tekalp   ZürichTürkiye, Koç University
Member of the Turkish Academy of Sciences

Member of Academia Europaea

IEEE Fellow

Professor, Koç University, Türkiye

Keynote Speech Title：Image Modeling: From handcrafted priors to foundation image models
Abstract：This talk unifies six decades of image modeling—evolving from handcrafted mathematical tractable priors to solve inverse problems in imaging to data-driven deep foundation image models for both image generation and solving inverse problems—under a common Bayesian framework that formulates image modeling as an implicit or explicit probabilistic characterization of a data manifold. Every image restoration or generative task fundamentally seeks to answer the same question: What makes an image look like a real image? In the Bayesian framework, this question is encapsulated by the prior probability density p(x), or its more optimization-friendly counterpart, the log-prior log p(x). While early works in solving inverse problems, such as Tikhonov, Total Variation (TV), and Non-local priors relied on explicit handcrafted definitions of log prior to model local smoothness, modern techniques, such as Plug-and-Play (PnP) priors and Denoising Diffusion models, are rooted in algorithmic uncoupling and implicit manifold learning that apply to both inverse problems and conditional image generation. By grounding imaging inverse problems and conditional image generation under a single mathematical engine: the gradient of the log-prior, or the Stein Score Function, we bridge the gap between signal processing and machine learning communities, turning what seems like parallel developments into a single, evolution of image modeling from handcrafted analytical models to models learned from big data, demonstrating that the pursuit of the 'denoising score' is the common thread connecting variational optimization (in MAP estimation) to posterior sampling (in modern foundation image models).
Biography：A. Murat Tekalp is a Professor at Koç University, Istanbul, Türkiye, where he served as Dean of Engineering from 2010 to 2013. His research interests include digital image and video processing, deep learning, video compression and streaming, multi-view and 3D video processing, and video networking.

Prof. Tekalp is an IEEE Fellow and a member of the Turkish Academy of Sciences and Academia Europaea. He received the TÜBİTAK Science Award, Türkiye's highest scientific award, in 2004 and the IEEE Turkey Section Lifetime Achievement Award in 2018.

He served as Editor-in-Chief of Signal Processing: Image Communication, published by Elsevier, from 1999 to 2010. He was also a member of the editorial boards of IEEE Signal Processing Magazine (2007–2010), Proceedings of the IEEE (2014–2019), and Wiley-IEEE Press (2018–2023).

Prof. Tekalp chaired the IEEE Signal Processing Society Technical Committee on Image and Multidimensional Signal Processing from 1996 to 1997. He was a founding member of the IEEE Technical Committee on Multimedia Signal Processing in 1997 and served on the committee until 2002.

He has held leadership roles at major international conferences, including Technical Program Co-Chair of IEEE ICASSP 2000, General Chair of IEEE ICIP 2002, and Technical Program Co-Chair of EUSIPCO 2005, IEEE ICIP 2020, and IEEE ICIP 2024.

He served on the European Research Council (ERC) evaluation panels for Advanced Grants (2009–2015), Starting Grants (2019–2025), and Synergy Grants (2022). He also served on the IEEE Signal Processing Society Fellow Evaluation Committee (2023–2025) and currently serves on its Nominations and Elections Committee.

Prof. Tekalp holds more than ten U.S. patents and is the author of the widely recognized textbook Digital Video Processing, first published by Prentice Hall in 1995, with its second edition released in 2015.

Moncef Gabbouj   Finland, Tampere University
Member of the Finnish Academy of Engineering Sciences

Member of the Finnish Academy of Science and Letters

Member of Academia Europaea

Professor of Information Technology, Tampere University, Finland

Keynote Speech Title：Blended Learning for High-Fidelity Volumetric CT Synthesis from Chest X-Rays
Abstract：Reconstructing volumetric Computed Tomography (CT) from a single 2D chest radiograph (CXR) is an ill-posed inverse problem, further complicated by the scarcity of paired CXR/CT training data. Prior approaches address this by training on Digitally Reconstructed Radiographs (DRRs), which are synthetic projections derived from CT volumes. However, the domain gap between DRRs and real CXRs limits generalization, often resulting in coarse or anatomically inconsistent reconstructions when applied to clinical images. To address this challenge, we introduce a Multi-Pass Multi-View Blended Learning framework to synthesize high-fidelity volumetric CT directly from real chest X-ray (CXR) images. The proposed approach progressively decomposes the synthesis task into two complementary learning stages. Stage 1 is an unsupervised CXR-to-DRR Domain Adaptation, while Stage 2 includes three passes, namely, a) supervised DRR-to-CT Transformation, b) unsupervised Multi-View Slice Refinement, followed by c) Progressive Transfer Learning (PTL). With such a blended learning paradigm, the proposed approach mitigates the synthetic-to-real domain gap while enhancing both the structural integrity and anatomical detail of the final output. On the LIDC-IDRI dataset, where paired DRR–CT ground truth is available for quantitative evaluation, the proposed method improves prior methods by up to 14% in PSNR and 7.6% in SSIM. The framework successfully generates structurally consistent and anatomically realistic high-fidelity CT volumes from real CXRs, marking a significant advancement toward clinical viability of CT reconstruction from standard radiographic images.
Biography：Moncef Gabbouj is a Professor of Information Technology in the Department of Computing Sciences at Tampere University, Finland. He served as an Academy of Finland Professor from 2011 to 2015. His research interests include artificial intelligence, machine learning, big data analytics, pattern recognition, multimedia content analysis and retrieval, nonlinear signal and image processing, voice conversion, and video processing and coding.

Prof. Gabbouj is an IEEE Fellow, a Fellow of the Asia-Pacific Artificial Intelligence Association (AAIA), and a member of Academia Europaea, the Finnish Academy of Science and Letters, and the Finnish Academy of Engineering Sciences.

He is a former Chair of the IEEE Circuits and Systems Society Technical Committee on Digital Signal Processing. He has served on the IEEE Signal Processing Society Fourier Award Committee and as Vice Chair of the IEEE Computer Society Harry Goode Award Committee. He has also served as an Associate Editor and Guest Editor for numerous IEEE and other international journals, and as a Distinguished Lecturer of the IEEE Circuits and Systems Society.

Prof. Gabbouj has held major leadership roles at international conferences, including serving as General Co-Chair of IEEE ISCAS 2019, IEEE ICIP 2020 and 2024, and IEEE ICME 2021.

He has served as Finland Site Director of the NSF-funded Industry–University Cooperative Research Centers for Visual and Decision Informatics (CVDI) and Big Learning (CBL). He also led the Artificial Intelligence Research Task Force of the Research Alliance on Autonomous Systems (RAAS), funded by Finland's Ministry of Economic Affairs and Employment.

Prof. Gabbouj has been an Honorary Professor at China University of Petroleum (East China), Qingdao, China, since 2023.