جستجوی استاد راهنمای دانشگاهی

Adway Girishمشاهده پروفایل

دانشگاهی

• Information Theory
• Coding Theory
• Machine Learning
• +۹ مورد دیگر

Adway Girish is a third-year Ph.D. candidate and Doctoral Assistant at the School of Computer and Communication Sciences (IC), École Polytechnique Fédérale de Lausanne (EPFL), Switzerland. He is affiliated with the Information Processing Group (IPG) and the Laboratory for the Theory of Information (LTHI), working under the supervision of Prof. Emre Telatar. He has also collaborated with leading researchers including Michael Gastpar, Hyeji Kim, and Shlomo Shamai. His educational background includes a B.Tech. in Electrical Engineering with honors and a minor in Mathematics from the Indian Institute of Technology Bombay (IITB), completed in 2022. Adway's research centers on information theory and its applications in communication, security, and machine learning. He is particularly interested in foundational aspects of information measures, entropy optimization, and the theoretical underpinnings of modern deep learning architectures such as transformers and large language models. His recent work explores rate-distortion frameworks for prompt compression, learning dynamics in transformers, and entropy-constrained communication channels. His research bridges theoretical rigor with practical implications in AI and communication systems. The trend in his recent publications shows a shift from classical signal processing and micro-Doppler analysis during his undergraduate years to advanced topics in information theory and machine learning during his Ph.D. His contributions are published in top-tier venues such as ISIT, NeurIPS, ICLR, and ICML workshops, indicating a strong trajectory in theoretical computer science and applied mathematics. Scientific Awards and Recognitions: ICLR 2025 Spotlight Paper (awarded to top 5% of accepted papers) Oral presentation at ICML 2024 Workshop on Theoretical Foundations of Foundation Models (selected as one of top 4 out of 58 submissions) Adway advises no students currently, as he is himself a doctoral candidate. However, he plays an active role in collaborative research projects involving multiple co-authors across institutions. He has not received specific mention of external grants in the provided text, but his position as a Doctoral Assistant at EPFL suggests institutional funding. His collaborations with renowned researchers suggest involvement in larger research initiatives and potential access to grant-supported projects. He is a core member of the Information Processing Group (IPG) at EPFL, a research lab focused on theoretical and applied aspects of information science, including coding, communication, learning, and data analysis. The group fosters interdisciplinary research and hosts regular seminars, candidacy reviews, and internal presentations, all of which Adway actively participates in.

Paolo Favaroمشاهده پروفایل

استاد

• Computer Vision
• Computational Photography
• Machine Learning
• +۴ مورد دیگر

Prof. Dr. Paolo Favaro is a full Professor and Head of the Computer Vision Group (CVG) at the University of Bern's Institute of Computer Science. He holds a Laurea (BSc+MSc) from the University of Padova, Italy, and M.Sc. and Ph.D. in Electrical Engineering from Washington University in St. Louis. His career includes postdoctoral roles at UCLA and Cambridge University, followed by work in medical imaging at Siemens Corporate Research. From 2006–2011, he was Lecturer and Reader at Heriot-Watt University, and Honorary Fellow at the University of Edinburgh before joining Bern in 2012. His research focuses on computer vision , computational photography , machine learning , and signal/image processing , with contributions to inverse problems and variational techniques. His work spans theoretical advancements and practical applications, such as video diffusion models, unsupervised segmentation, and image restoration. He is an active member of the IEEE Society. Recent research trends emphasize generative models (e.g., diffusion models, video generation) and self-supervised learning , addressing challenges in real-world image/video denoising and sparsity-driven methods. His articles explore topics like Kalman-based optimization, multimodal world models, and spatio-temporal representation learning. Prof. Favaro leads the CVG, fostering interdisciplinary research at the intersection of computer vision and AI. His lab develops cutting-edge solutions for 3D reconstruction, medical imaging, and autonomous systems.

Anastasiia Shynkarenkoمشاهده پروفایل

پژوهشگر

• Seismology
• Geotechnical Engineering
• Natural Hazards
• +۳ مورد دیگر

Dr. Anastasiia Shynkarenko is a Researcher at the Swiss Seismological Service (SED) within ETH Zürich's School of Architecture, Civil and Environmental Engineering. Her work focuses on seismic risk assessment, subaqueous slope stability, and lake tsunami hazards in Swiss perialpine lakes. She specializes in integrating geophysical and geotechnical methods to investigate sediment properties and failure mechanisms using advanced techniques like Ocean Bottom Seismometers (OBS), ambient vibration analysis, and in-situ testing (CPTu). Key projects include the comprehensive study of Lake Lucerne's submerged slopes, addressing slope stability through interdisciplinary approaches. Her research combines seismological data inversion, spectral analysis, and numerical modeling to evaluate subsurface structures and potential hazards. Recent contributions include developing dynamic earthquake risk frameworks and advancing methodologies for extracting phase velocities from OBS arrays to characterize sedimentary sites. While no formal student advising or grant information is explicitly stated, her work has been published in leading geophysics and engineering journals. Collaborations involve both academic and governmental entities in Switzerland focused on environmental and geohazard mitigation strategies.

Christos Dimitrakakisمشاهده پروفایل

استاد

• Reinforcement Learning
• Reverse Reinforcement Learning
• Preference Elicitation
• +۱۱ مورد دیگر

Christos Dimitrakakis is a Full Professor of Data Science at the University of Neuchâtel, Switzerland, where he is affiliated with the Institute of Computer Science within the Faculty of Science. His research focuses on artificial intelligence, particularly reinforcement learning, fairness, and algorithmic privacy. He has held academic positions at several prestigious institutions including the University of Oslo, Chalmers University of Technology, and the University of Lille. His educational background includes: 1997: BEng from the University of Manchester 1998: MSc from the University of Essex 2006: PhD from IDIAP/EPFL under the supervision of Samy Bengio Professor Dimitrakakis specializes in reinforcement learning and decision making under uncertainty, with particular expertise in reverse reinforcement learning, preference elicitation, and human-AI collaboration. His research bridges theoretical foundations with practical applications, especially in autonomous systems and privacy-preserving AI. He has made significant contributions to understanding representations of uncertainty, differential privacy, and group fairness in algorithmic decision-making. His work often combines Bayesian inference with reinforcement learning frameworks to develop robust and fair AI systems that can operate effectively in uncertain environments. His recent publications demonstrate a strong focus on theoretical foundations of reinforcement learning while addressing practical challenges in fairness, privacy, and decision making. The research shows an evolving trajectory from core reinforcement learning techniques toward applications in fair and transparent AI systems, with increasing emphasis on real-world constraints and ethical considerations. His work spans multiple domains including game theory, matching markets, and bandit algorithms, reflecting an interdisciplinary approach to AI research. Notable scientific achievements include: Best Paper Award at Cooperative-AI@NeurIPS 2021 for "Interactive Inverse Reinforcement Learning for Cooperative Games" Student paper award at EAAMO'22 for "On Meritocracy in Optimal Set Selection" Professor Dimitrakakis actively mentors PhD students and has supervised numerous researchers who have gone on to academic and industry positions. His current research group at the University of Neuchâtel includes five PhD students working on reinforcement learning, fairness, and uncertainty in AI systems. He has secured funding for projects including a national initiative to develop intelligent bird deterrent systems for crop protection, demonstrating the practical impact of his research. His group collaborates with industry partners such as Agroscope, and he maintains academic connections with institutions worldwide. His research group at the University of Neuchâtel is actively working on interdisciplinary projects, most notably the development of an "intelligent scarecrow" system for bird pest control in agriculture. This project combines computer vision, reinforcement learning, and behavioral modeling to create adaptive AI systems that can minimize crop damage while accounting for bird behavior patterns. The group maintains expertise in both theoretical foundations and practical implementations of AI systems.

Raoul de Charetteمشاهده پروفایل

استاد

• Visual Scene Understanding
• Computer Vision
• 3D Reconstruction
• +۳ مورد دیگر

Raoul de Charette is a Research Director in computer vision at Inria Paris, leading the Astra-Vision group within the ASTRA team. His academic journey includes a PhD from Mines Paris (2012) and Habilitation (HDR) in 2022, with research stints at Carnegie Mellon University (2011), Mines Paris (2013), and the University of Makedonia (2014). His educational background comprises: PhD from Mines Paris (2012) Habilitation (HDR) (2022) De Charette's research centers on robust and interpretable visual scene understanding , spanning 3D scene reconstruction, domain adaptation, material recognition, and physics-grounded vision foundation models. His work integrates physical principles and synthetic data to enhance model robustness in real-world scenarios like autonomous driving and urban environments. Key contributions include uncertainty-aware 3D scene completion (PaSCo), material extraction from single images (Material Palette), and prompt-driven domain adaptation (PODA). Recent publications reveal a strategic shift toward vision-language integration, material-centric scene understanding, and foundation models that minimize labeled data dependency. His group pioneers physics-informed approaches to improve interpretability and resilience against environmental challenges like adverse weather conditions. Key scientific recognition includes: Best Paper Honorable Mention at EGSR 2025 for MatSwap ELLIS Membership PR[AI]RIE-PSAI Fellowship De Charette actively mentors four PhD students—Fatima Balde, Mohammad Fahes, Ivan Lopes, and Tetiana Martyniuk—often in industry collaborations with Valeo.ai and Kyutai. He secures funding through fellowships and industry partnerships, regularly opening PhD positions (including a 2025 opening for Physics-Grounded Vision Foundation Models). As an area chair for CVPR, ECCV, WACV, and IROS, he shapes the field through conference leadership and co-organizing initiatives like the African Computer Vision Summer School. He directs the Astra-Vision group within Inria Paris' ASTRA team, driving interdisciplinary research at the intersection of computer vision, machine learning, and physics-based modeling for real-world deployment in robotics and intelligent transportation systems.

Martin Zachمشاهده پروفایل

پژوهشگر ارشد

• Biomedical Imaging
• Inverse Problems
• MRI Reconstruction
• +۲ مورد دیگر

Martin Zach is a postdoctoral researcher at the Center for Biomedical Imaging (CIBM), EPFL , specializing in inverse problems in biomedical imaging. He joined the Mathematical Imaging Section under the supervision of Prof. Michael Unser in September 2024. PhD : Graz University of Technology (2024), advised by Thomas Pock Martin’s research bridges model-based reconstructions and data-driven approaches in imaging, with a focus on MRI reconstruction and diffusion models . His work explores regularization techniques, energy-based priors, and probabilistic modeling. Recent publications highlight advancements in Gaussian mixture models , non-linear inversion , and Langevin sampling for biomedical imaging. His Google Scholar profile reveals a strong focus on inverse problems (2020–2025), with applications in MRI , CT , and quantitative phase imaging . Key methodologies include diffusion models , generative priors , and regularization algorithms , spanning disciplines from machine learning to computational biology . Martin’s current role at EPFL involves collaborative research in mathematical imaging , with affiliations to the BioMedical Imaging (BIG) Group in Lausanne, Switzerland.

Valery Vishnevskiyمشاهده پروفایل

مدرس

• Biomedical Imaging
• Magnetic Resonance Imaging (MRI)
• Ultrasound Imaging
• +۳ مورد دیگر

Dr. Valery Vishnevskiy is a Lecturer at the Department of Information Technology and Electrical Engineering at ETH Zürich , affiliated with the Biomedical Imaging Group . His research focuses on advanced medical imaging techniques, particularly in cardiac MRI , ultrasound , and machine learning applications . He contributes to projects involving 4D flow MRI reconstruction , speed-of-sound imaging , and motion-corrected diffusion tensor analysis . Primary Affiliation : ETH Zürich, Department of Information Technology and Electrical Engineering Research Interests : Biomedical Imaging, MRI Reconstruction, Ultrasound Physics, Cardiac Imaging, Deep Learning His recent work emphasizes self-supervised learning for medical imaging, including the development of FlowMRI-Net for accelerated 4D flow MRI and Speed-of-Sound Imaging using diverging waves. Publications span hyperpolarized 13C metabolic imaging , viscoelasticity reconstruction , and probabilistic sampling optimization . Dr. Vishnevskiy's contributions to deformable image registration include methods for handling sliding interfaces in abdominal and cardiac imaging, with applications in respiratory motion compensation and tissue ablation monitoring . He also explores mathematical models for enzyme activity analysis and behavioral pattern detection.

Andrea Zuninoمشاهده پروفایل

مدرس

• Inverse Problems
• Seismology
• Computational Geoscience
• +۳ مورد دیگر

Dr. Andrea Zunino is a Lecturer at the Institute of Geophysics, ETH Zurich, within the Department of Earth and Planetary Sciences. His research focuses on developing methodologies to infer subsurface Earth properties using probabilistic approaches, including Monte Carlo methods. Key areas include inverse problems, seismology, and computational geoscience. He previously collaborated with Klaus Mosegaard on probabilistic inverse problems and geostatistics. Current work involves ambient noise seismic tomography and Hamiltonian Monte Carlo (HMC) strategies for geophysical problems. Research interests emphasize nonlinear inverse problem solutions, particularly through HMC frameworks like HMCLab, an open-source tool for diverse geophysical applications. His work integrates adjoint methods for efficient sampling and uncertainty quantification. Notable projects include borehole fiber-optic seismology in Greenland, aeromagnetic studies of Antarctic volcanic fields, and joint gravity-magnetic inversion using HMC. Publications highlight contributions to geophysical inversion techniques, seismic data analysis, and computational frameworks. He leads the Seismology and Wave Physics group, advancing probabilistic methodologies for Earth structure imaging and subsurface characterization.

Hans-Andrea Loeligerمشاهده پروفایل

استاد

• Signal Processing
• Information Theory
• Communications
• +۶ مورد دیگر

Prof. Hans-Andrea Loeliger is a Full Professor at ETH Zurich's Department of Information Technology and Electrical Engineering and serves as Deputy Head of the Signal and Information Processing Laboratory. With a career spanning over two decades at ETH Zurich since 2000, he has established himself as a leading researcher in signal processing, information theory, and related fields. His work bridges theoretical foundations with practical applications in communications, electronics, and machine learning. Loeliger's research interests encompass a broad spectrum of topics including signal processing, information theory, communications, system theory, electronics, machine learning, quantum systems, error correcting codes, and neural computation. His work on factor graphs and message passing algorithms has been particularly influential, providing a unifying framework for various signal processing techniques. His recent publications demonstrate continued innovation in areas such as NUV priors, control-bounded analog-to-digital conversion, and neural network applications. His publication record shows consistent high-impact contributions across multiple disciplines, with recent work focusing on the intersection of statistical signal processing, machine learning, and circuit design. The trend in his research demonstrates an evolution from foundational work in factor graphs and information theory toward increasingly sophisticated applications in machine learning, neural computation, and practical circuit implementations. Fellow of the IEEE Loeliger has supervised numerous PhD students and master's candidates through ETH Zurich's Signal and Information Processing Lab, though specific student names aren't listed in the provided text. His teaching responsibilities include courses such as Discrete-Time and Statistical Signal Processing, Electronic Circuits & Signals Exploration Laboratory, and Introduction to Estimation and Machine Learning. His research has been supported by various grants enabling the development of novel signal processing techniques and their implementation in practical systems. The Signal and Information Processing Laboratory under his leadership serves as a hub for interdisciplinary research connecting theoretical signal processing with applications in communications, imaging, and neural systems. The lab maintains strong connections with both academic and industrial partners, facilitating the translation of theoretical advances into practical implementations.

AmirEhsan Khorashadizadehمشاهده پروفایل

پژوهشگر ارشد

• Machine Learning
• Computational Imaging
• Inverse Problems
• +۷ مورد دیگر

AmirEhsan Khorashadizadeh is a postdoctoral researcher affiliated with the Paul Scherrer Institute (PSI) and École Polytechnique Fédérale de Lausanne (EPFL) , working in the Computational X-ray Imaging group under Prof. Manuel Guizar Sicairos. His research integrates deep learning and computational imaging , focusing on physics-informed neural networks for applications in tomography , inverse scattering , and cosmological imaging . He collaborates with the Swiss Data Science Center (SDSC) on the CHIP project 'Machine-Learning-assisted Ptychographic nanotomography.' His work spans multiple domains, including medical imaging , Bayesian imaging , and astrophysical signal processing . Publications highlight advancements in scalable image reconstruction , injective flow-based models , and implicit neural representations , with applications in X-ray and cosmological imaging . Scientific Awards Promotion of Young Talent grant from University of Basel for 9-month visiting scholar at University College London (UCL) Top 25 most downloaded paper in IEEE Transactions on Computational Imaging (TCI) from Sept. 2022–Sept.2023

• Machine Learning
• High Energy Physics
• Computational Physics
• +۳ مورد دیگر

Gregor Kasieczka is a Professor at the University of Hamburg, affiliated with the Faculty of Mathematics and Natural Sciences and the Department of Physics. His research focuses on applying machine learning techniques to high-energy physics, particularly for LHC event generation, detector simulations, and jet reconstruction. He collaborates extensively with physicists and computer scientists globally. Research Trends: His recent work explores equivariant neural networks for particle jets, invertible models for detector simulations, and uncertainty quantification in ML-driven calibration. He also investigates generative models (GANs) to amplify event samples and improve resolution beyond traditional methods. Collaborations: Gregor works with prominent researchers like Tilman Plehn, Benjamin Nachman, and Jesse Thaler, contributing to interdisciplinary advancements in ML applications for particle physics. Labs & Teams: He leads the Gruppe Kasieczka at the University of Hamburg, specializing in machine learning for experimental particle physics and detector-level simulations.

Pier Luigi Dragottiمشاهده پروفایل

استاد

• Signal Processing
• Wavelet Theory
• Sampling Theory
• +۷ مورد دیگر

Prof. Pier Luigi Dragotti is a Professor of Signal Processing in the Department of Electrical and Electronic Engineering at Imperial College London, Faculty of Engineering. He leads the Communications, Signal Processing, and Control (CSP) research group and is actively engaged in both theoretical and applied research in signal processing and computational imaging. His research focuses on wavelet theory, sampling theory, sparse signal processing, computational imaging, and data-driven image processing . He integrates classical signal processing frameworks with modern deep learning techniques, particularly in solving inverse problems and image restoration. His work bridges model-based algorithms with data-driven approaches, enabling breakthroughs in fields such as art conservation and neuroscience imaging. The recent publications highlight a strong trend towards invertible neural networks, physics-informed deep learning, and hybrid model-data methods for image restoration, microscopy, and cultural heritage analysis. His team develops algorithms that are not only accurate but also interpretable and grounded in physical models. Scientific Awards and Recognitions: Editor-in-Chief of IEEE Transactions on Signal Processing (2018–2020), Certificate of Merit Supervised students winning: Eurasip 3M Thesis Competition, IEEE MMSP Best Paper Award, Outstanding PhD Thesis Awards (2020–2022), Ivor Tupper Prize Student Adam won Gold at the Commonwealth Games Advising and Grants: Prof. Dragotti has successfully supervised numerous PhD and MSc students, many of whom have gone on to achieve significant recognition. His group has secured funding that supports interdisciplinary research in imaging science, and several of his former students have launched innovative projects, including a start-up company. He fosters a dynamic research environment with strong industry and cross-institutional collaborations. Labs and Research Groups: He is a key member of the CSP (Communications, Signal Processing, and Control) group at Imperial College London, where he leads a vibrant team working on cutting-edge signal processing challenges. The group is known for its contributions to sampling theory, wavelets, and the integration of deep learning in imaging systems.

• Biomedical Imaging
• Applied Functional Analysis
• Sampling Theory
• +۴ مورد دیگر

Michael Unser is a Full Professor at EPFL’s School of Engineering and Academic Director of EPFL’s Center for Imaging in Lausanne, Switzerland. His research focuses on biomedical imaging, applied functional analysis, sampling theory, wavelets, splines, and computational bioimaging. Current Position: Full Professor, EPFL School of Engineering Academic Role: Director of EPFL Center for Imaging Key Research Areas: Inverse problems, sparse stochastic processes, machine learning for imaging Dr. Unser’s research bridges mathematical theory with practical biomedical imaging applications. He has pioneered the use of splines for image processing, developed advanced regularization frameworks for inverse problems, and contributed to computational bioimaging techniques like SMLM and optoacoustic tomography. His recent work explores connections between deep learning and classical kernel methods through variational formulations. He has published over 400 journal papers and authored the seminal book on sparse stochastic processes. His scientific contributions have been recognized with multiple awards including IEEE and EURASIP Technical Achievement Awards, five IEEE-SPS Best Paper Awards, and three consecutive ERC Advanced Grants (FUN-SP, GlobalBioIm, FunLearn). He has served on editorial boards for leading journals and founded the IEEE SPS Bio Imaging and Signal Processing Technical Committee.

Tailin Wuمشاهده پروفایل

استادیار

• Generative AI for scientific simulation
• AI Agents for scientific discovery
• Representation learning with graph neural networks
• +۳ مورد دیگر

Tailin Wu is an Assistant Professor at Westlake University in the Department of AI within the School of Engineering . He leads the AI for Scientific Simulation and Discovery Lab , focusing on integrating machine learning with fundamental scientific domains. Previously, he conducted postdoctoral research at Stanford Computer Science (2020-2023) under Jure Leskovec , earned his PhD in Physics from MIT (2019) under advisors Isaac Chuang and Max Tegmark , and completed his BSc in Physics at Peking University (2012). Research Interests : Generative AI for scientific simulation (diffusion models, flow matching), AI agents for automated discovery (neural-symbolic integration), representation learning with graph models and information theory His work spans physics-informed machine learning, with applications in fluid dynamics, energy systems, and life sciences. Key contributions include developing methods for safe PDE control with conformal prediction and creating interpretable theory-learning frameworks through the AI Physicist paradigm. Scientific Awards : Best Poster Award, ICML Time Series Workshop (2019) Outstanding Postdoc, Westlake University (2024) Nominated for Outstanding Youth Paper Award, China Embodied AI Conference (2025) As an educator, he teaches Frontiers in Computer Science and Technology at Westlake University and has guest-lectured at Caltech and MIT. His lab actively recruits postdocs, PhD students, and interns.

David Ginsbourgerمشاهده پروفایل

استاد

• Statistics
• Machine Learning
• Uncertainty Quantification
• +۷ مورد دیگر

David Ginsbourger is a Professor and Head of Research Group at the Institute of Mathematical Statistics and Actuarial Science (IMSV) within the University of Bern, Switzerland. He maintains dual affiliations through his role at IMSV and as a member of the Multidisciplinary Center for Infectious Diseases (MCID), reflecting interdisciplinary engagement across statistical methodology and applied domains. His research program centers on advanced statistical methodologies with emphases on Gaussian process modeling, uncertainty quantification, and experimental design for computer experiments. Key contributions include novel kernel constructions for equivariant systems, sequential design strategies for excursion set estimation, and efficient computational frameworks for spatial distributional modeling. His work bridges theoretical statistics with practical applications in agriculture, chemoinformatics, environmental science, and risk assessment, demonstrating consistent innovation in handling complex prediction problems under uncertainty. Analysis of his 15 most recent publications (2024-2025) reveals persistent methodological development in Gaussian process theory alongside expanding application domains. Recurring themes include integration-free kernel design for structured data, rare event probability estimation, and multivariate forecast calibration. His research exhibits strong continuity in addressing computational challenges for large-scale inverse problems while increasingly incorporating domain-specific constraints from fields like molecular chemistry and agricultural science. Ginsbourger leads a dedicated research group at IMSV focused on advancing statistical frameworks for computer experiments and uncertainty quantification. The group maintains active collaborations across disciplines, particularly evident in recent work connecting statistical methodology to infectious disease modeling through MCID affiliations and agricultural optimization projects.