Find Academic Supervisors

• Computer Vision
• Computer Graphics
• Machine Learning
• +11 more

Andrea Tagliasacchi is an Associate Professor in the School of Computing Science at Simon Fraser University (SFU), where he holds the Visual Computing Research Chair. He is also a part-time (20%) Staff Research Scientist at Google DeepMind in Toronto and holds an associate professor (status only) appointment in the Department of Computer Science at the University of Toronto. Education: PhD in Computing Science – Simon Fraser University (NSERC Alexander Graham Bell Fellow) Postdoctoral Research – École Polytechnique Fédérale de Lausanne (EPFL) MSc in Computer Science – Politecnico di Milano (Gold Medalist) His research lies at the intersection of computer vision, computer graphics, and machine learning, with a focus on 3D visual perception. Key areas include neural radiance fields (NeRF), 3D Gaussian splatting, inverse rendering, and geometric deep learning, with applications in robotics, augmented reality, and autonomous systems. His work emphasizes robust and efficient scene understanding and reconstruction from visual data. His recent publications, appearing in top venues like CVPR, SIGGRAPH, NeurIPS, and ECCV, demonstrate a strong emphasis on neural fields, 3D reconstruction, and generative modeling. Trends include improving rendering efficiency, enhancing robustness to noise and distractors, and enabling controllable and 3D-aware generation. His group has made significant contributions to Gaussian splatting, NeRF optimization, and diffusion-based 3D/4D synthesis. Scientific Awards: 2024 CVPR Best Paper Award (Honorable Mention) 2020 CVPR Best Student Paper Award 2015 SGP Best Paper Award NSERC Alexander Graham Bell Canada Graduate Scholarship MITACS Best Paper Award (SIGGRAPH Asia 2009) NSF Best Poster Award (SGP 2012) He has advised numerous PhD and MSc students, many of whom are now researchers at leading institutions and companies. His research has been supported through collaborations with Google, Intel, and academic partners. He serves the community as a Senior Area Chair for CVPR 2025, Associate Editor for IEEE TPAMI (2024–2026), Guest Editor for IEEE TPAMI on 3D GenAI, and Program Chair for 3DV 2024. He leads a vibrant research lab at SFU focused on pushing the boundaries of 3D scene understanding with machine learning.

Agnes DesolneuxView profile

Professor

• Image Analysis
• Statistical Methods
• Computer Vision
• +7 more

Agnes Desolneux is a CNRS Research Director at the Borelli Centre (formerly CMLA) and a Professor attached to the Mathematics Department at ENS Paris-Saclay. Education: PhD in Applied Mathematics (2000) from ENS Cachan Habilitation in Applied Mathematics (2010) from Université Paris Descartes Her research focuses on image analysis via statistical methods , particularly a contrario approaches, image restoration, texture synthesis, Determinantal Point Processes (DPP), optimal transport, Gaussian mixtures, geometry of random field excursions, shot-noise models, and mathematical modeling of visual perception through Gestalt theory. The articles extracted reflect her expertise in applied mathematics and computer vision , with recent works (2025-2020) on optimal transport algorithms, DPP applications, multiscale texture analysis, and stochastic modeling in medical imaging. Keywords span machine learning, probability theory, medical imaging, and computer vision . She has no listed scientific awards but has authored influential works including the book From Gestalt Theory to Image Analysis: A Probabilistic Approach (Springer, 2008) and Pattern Theory: the stochastic analysis of real-world signals (AK Peters, 2010).

Mathias NiepertView profile

Professor

• Artificial Intelligence
• Machine Learning
• Graph Neural Networks
• +5 more

Mathias Niepert is a Professor at the Institute for Artificial Intelligence within the Faculty of Computer Science, Electrical Engineering and Information Technology at the University of Stuttgart. His research focuses on advancing machine learning techniques with applications in scientific computing, graph neural networks, and medical imaging. He is particularly known for contributions to physics-informed neural networks, equivariant models, and graph learning frameworks. Key research areas include: Scientific Machine Learning for PDEs and molecular modeling Graph neural networks and their theoretical limitations Medical vision-language models and multimodal learning Efficient neural network architectures (transformers, FNOs) Domain knowledge integration in deep learning His work often bridges theoretical foundations with practical applications, as evidenced by extensive publications (2018–2025) on topics like adaptive message passing, equivariant networks, and medical imaging systems. He has contributed to benchmark development through initiatives like PDEBench and pioneered methods for equivariant diffusion models and molecular representation learning. His current projects emphasize: Improving generalization in Fourier Neural Operators Addressing oversmoothing in graph networks Combining physics principles with neural architectures Medical AI applications through multimodal fusion

Jan PetersView profile

Professor

• Robot Learning
• Reinforcement Learning
• Imitation Learning
• +5 more

Jan Peters is a full professor (W3) at the Computer Science Department of Technische Universität Darmstadt and serves as the department head of the Systems AI for Robot Learning (SAIROL) at the German Research Center for AI (DFKI) . He is also a founding faculty member of the Hessian Centre for Artificial Intelligence . Peters holds a Ph.D. in Computer Science from the University of Southern California (2007) and dual master’s degrees in Computer Science and Electrical Engineering from USC and TU Munich respectively. Research Themes : Robot Learning, Reinforcement Learning, Imitation Learning, Tactile Sensing, Human-Robot Interaction, and Safe AI. Recent Article Trends : Focus on deep reinforcement learning (Iterated Q-Networks, Adaptive Q-Networks), safe robot foundation models , tactile-enhanced imitation learning , and physics-informed machine learning . Scientific Recognition : Recipient of the Dick Volz Best PhD Thesis Award , ERC Starting Grant , IEEE Fellow , and Amazon Research Award . Leadership : Founder of the IEEE RAS Technical Committee on Robot Learning and editor for journals including Autonomous Robots and IEEE Transactions on Robotics .

Felix KrahmerView profile

Assistant Professor

• Compressed Sensing
• Signal Processing
• Optimization
• +4 more

Prof. Felix Krahmer is a TUM Tenure Track Assistant Professor of Optimization and Data Analysis at the Department of Mathematics, Technische Universität München (TUM), part of the School of Computation, Information and Technology. Previously, he held a Junior Professorship (W1) for Mathematical Data Analysis at the University of Göttingen (2012–2015). His research focuses on mathematical foundations of data science, including compressed sensing, signal quantization, uncertainty quantification, and optimization algorithms. He has led an Emmy Noether research group and contributed to the ZeMat interdisciplinary project. Education : PhD in Mathematics (2009), New York University, advisors: Percy Deift & Sinan Güntürk MSc in Mathematics (2007), New York University BSc in Mathematics (2004), Jacobs University Bremen Research Interests : Krahmer’s work bridges theoretical mathematics and applied data science. Key areas include compressed sensing algorithms, high-dimensional signal reconstruction, quantization theory for analog-to-digital conversion, and optimization methods for inverse problems. He also explores applications in imaging (e.g., MRI reconstruction) and stochastic processes. Key Contributions : His research on phase retrieval, sigma-delta quantization, and compressed sensing recovery has advanced signal processing techniques. Recent efforts focus on uncertainty quantification for high-dimensional inverse problems and the mathematical analysis of consensus-based optimization. Grants & Awards : Emmy Noether Research Group Grant (DFG) Funding from BMBF (ZeMat project) Teaching & Outreach : Krahmer has taught advanced courses on compressed sensing, functional analysis, and random matrix theory. He co-organized workshops on probabilistic techniques and clinical risk prediction, emphasizing interdisciplinary collaboration. Labs/Teams : Member of the TUM Data Science Research Group, focusing on optimization, imaging, and uncertainty quantification. Active in the Munich Center for Quantum Science and Technology (MCQST).

• Climate Change Adaptation in Agriculture
• Land Use Modeling
• Biodiversity Conservation
• +3 more

Thomas Berger is a Professor at the University of Hohenheim , affiliated with the Faculty of Agricultural Sciences and leading the Department of Economics of Land Use . He also contributes to the Computational Science Hub and Hohenheim Tropics initiatives. Focus Areas: Climate change adaptation, land-use modeling, biodiversity-productivity trade-offs, agent-based simulation, and machine learning in agricultural systems. Key Projects: Simulation frameworks for smallholder resilience in Ethiopia, bioeconomic modeling in the Amazon, and hybrid intelligence applications in European agricultural policy. Recent Publications: 2025 study on climate change effects on insecticide reduction in Germany, 2024 work on reconciling biodiversity with productivity via hybrid models, and 2023 methodological contributions to surrogate modeling and seasonal forecast integration. Research Trends: Interdisciplinary integration of climate science, agricultural economics, and computational modeling, with increasing emphasis on AI-assisted decision support systems and sustainability policy validation. Teaching & Outreach: Offers Agricultural Economics seminars and Hohenheim Tropics discussions, requiring advance email registration for office hours.

Christian WaldView profile

Research Fellow

• Generative Modeling
• Flow Matching
• Stochastic Processes
• +3 more

Christian Wald is a Post-doctoral researcher at Technical University Berlin working under Professor Gabriele Steidl, focusing on generative modeling, flow matching, and stochastic processes in machine learning. His research bridges theoretical probability with practical medical imaging applications, particularly in MRI reconstruction and analysis. He completed his PhD at Humboldt University of Berlin in 2017 with a thesis on p-adic quantum groups. His academic journey transitioned from pure mathematics to interdisciplinary machine learning research, reflecting his versatile expertise. Wald's primary research explores generative models through the lens of optimal transport and flow matching, with significant contributions to Wasserstein geometry and conditional distance metrics. His work frequently integrates stochastic processes to enhance medical image reconstruction, demonstrating strong cross-disciplinary impact in both theoretical machine learning and clinical applications. Recent publications highlight innovations in sliced MMD flows, Bayesian OT methods, and uncertainty-aware medical image analysis. Analysis of his 15 most recent publications (2019-2025) reveals a consistent trajectory toward unifying geometric probability with deep learning. Key themes include flow-based generative modeling for medical time-series data, optimal transport applications in image reconstruction, and novel kernel methods for distribution matching. His work spans both foundational theory (e.g., Fisher-Rao curves) and high-impact medical applications (e.g., coronary calcium scoring). No specific scientific awards are documented in the provided text, though his publications appear in prestigious venues including ICLR, IEEE TMI, and Physics in Medicine & Biology. Wald maintains extensive collaborations with the medical imaging group at Technical University Berlin, particularly with Andreas Kofler and Gabriele Steidl. His co-authored works demonstrate consistent contributions to MRI reconstruction pipelines and segmentation frameworks, though no formal advising roles or grant leadership are indicated. Current projects focus on uncertainty quantification in active learning for medical image segmentation. He operates within Gabriele Steidl's research group at Technical University Berlin, which specializes in mathematical imaging and machine learning. The team combines expertise in optimization, probability theory, and deep learning to solve medical imaging challenges, with Wald contributing core algorithmic innovations in generative modeling and stochastic reconstruction.

Marco CuturiView profile

Professor

• Machine Learning
• Optimal Transport
• Optimization
• +4 more

Marco Cuturi is a Research Scientist at Apple ML Research in Paris and Professor of Statistics at CREST-ENSAE, Institut Polytechnique de Paris. His work bridges machine learning , optimal transport , and optimization , with applications in time-series analysis , kernels , and multiresolution methods . He has held academic roles at Kyoto University and Princeton University, and previously worked in the financial industry. Research Interests: Optimal transport theory and computational methods Kernel design for structured data and histograms Time-series alignment and soft-DTW Entropic regularization in optimization Applications to computer vision and genomics Teaching: Cuturi has taught courses on linear optimization at Princeton, geometric methods in machine learning at Kyoto, and scientific English. He has also organized machine learning summer schools in Kyoto, Les Houches, and other international venues. Recent Trends: His 2024-2025 publications focus on entropic optimal transport solvers, disentangled representation learning via Gromov-Monge gaps, and applications to text-to-image diffusion models. Collaborative work with institutions like Google Research, MIT, and University of Tokyo highlights his interdisciplinary impact.

Constantin A. RothkopfView profile

Professor

• Perception and Action
• Computational Neuroscience
• Machine Learning
• +3 more

Prof. Constantin A. Rothkopf is a W3 Professor at the Department of Psychology, Technische Universität Darmstadt, and a secondary member of the Department of Computer Science. He serves as Founding Director of the Centre for Cognitive Science and founding member of the Hessisches Zentrum für Künstliche Intelligenz (hessian.ai). He is also part of the European Laboratory for Learning and Intelligent Systems (ELLIS) and the DAAD Konrad Zuse Schools of Excellence in Artificial Intelligence (ELIZA). His research focuses on the interplay between perception and action, using computational models and experimental studies in humans. Current work includes eye-tracking studies in naturalistic environments, inverse optimal control models, and developing algorithms for virtual agents. Education: Ph.D. in Neuroscience and Informatics from the University of Rochester, followed by postdoctoral research at Frankfurt Institute for Advanced Studies (FIAS). He has held visiting professorships at Central European University (2017) and Columbia University (2023). Awards include an ERC Consolidator Grant (2022) and SCENE Project Funding (2025). Research Interests: Active vision, decision-making under uncertainty, sensorimotor control, and computational modeling. Key themes include how humans use sensory input to form beliefs, make decisions, and act in dynamic environments. Grants/Awards: ERC Consolidator Grant (2022), SCENE Funding (2025) Labs/Teams: Centre for Cognitive Science, hessian.ai, ELLIS Unit Darmstadt

Jalal EtesamiView profile

Assistant Professor

• Machine Learning
• Statistical Decision Making
• Causal Inference
• +4 more

Prof. Jalal Etesami is an Assistant Professor in the Department of Computer Science at Technical University of Munich (TUM), leading the Decision Sciences & Systems group. He holds a Ph.D. in Industrial and Systems Engineering from the University of Illinois at Urbana-Champaign and was a Postdoctoral Fellow at EPFL in Switzerland. His research focuses on machine learning, causal inference, multi-agent systems, and game theory, with applications to systemic risk modeling and market design. He teaches advanced courses such as Causal Inference in Time Series , Algorithmic Game Theory , and Optimization, Learning, and Market Design . Notable contributions include work on causal structure learning, stochastic optimization, and non-Gaussian causal models. Recent research explores causal effect identification under confounding, neural networks for market analysis, and optimal experiment design. Prof. Etesami’s work appears in top venues like NeurIPS, AAAI, and IEEE journals. He actively contributes to the academic community, organizing seminars and workshops on topics ranging from causal reasoning to computational social choice.

Stefano NoventaView profile

Research Fellow

• Mathematical Psychology
• Psychometrics
• Psychophysics
• +7 more

Stefano Noventa is a Research Fellow at the Methods Center, Department of Social Sciences, Faculty of Economics and Social Sciences, University of Tübingen. He has held multiple postdoctoral positions at the University of Tübingen and previously at the University of Verona and the University of Padova. Education: Ph.D. in Cognitive Psychology, University of Padova (2011) M.Sc. in Physics, University of Padova (2006) Studies in Physics, University of Padova (1999–2006) International Visiting Graduate Student, University of Toronto (2009, 2010) Dr. Noventa's research lies at the intersection of mathematical psychology, psychometrics, and psychophysics, with a focus on developing and unifying quantitative models of human cognition and assessment. His work integrates Item Response Theory (IRT) and Knowledge Space Theory (KST) to create more robust frameworks for educational and psychological measurement. He investigates latent variable models, probabilistic knowledge structures, and the identifiability of complex psychometric models, often applying these to domains such as education, organizational psychology, and entrepreneurship. His recent publications (2020–2024) demonstrate a strong trend toward theoretical integration, particularly in bridging cognitive diagnosis models with traditional psychometric frameworks. The articles emphasize mathematical rigor, model generalization, and empirical validation, with applications in both cognitive science and applied psychology. Topics include the unification of assessment models, parameter estimation under local dependence, and the modeling of intuitive physical reasoning. Scientific Awards: No awards or honors listed in the provided text. Dr. Noventa has not been explicitly mentioned as an advisor to students, but he has served as a corresponding author and collaborator on multiple research projects, indicating a leadership role in research teams. He has been involved in a DFG-funded project (GLI NON-NORM) since 2019, suggesting active grant participation. His work is highly collaborative, involving researchers from Germany, Italy, Austria, and Canada. Labs and Research Groups: Methods Center, University of Tübingen Hector Institute of Education Science and Psychology, University of Tübingen Center of Assessment, University of Verona Department of General Psychology, University of Padova

Max FathiView profile

Professor

• Optimal Transport
• Probability Theory
• Statistical Physics
• +9 more

Max Fathi is a Professor of Mathematics at Université Paris Cité, affiliated with the Laboratoire Jacques-Louis Lions (LJLL) and Laboratoire de Probabilités, Statistique et Modélisation (LPSM). He concurrently holds a part-time teaching position at the Department of Mathematics and Applications (DMA) at École Normale Supérieure (ENS). Since 2023, he has been a member of the Institut Universitaire de France (IUF), a prestigious national research fellowship in France. He completed his PhD in 2013 at Université Pierre et Marie Curie under Cédric Villani, followed by a postdoctoral position at the University of California, Berkeley with Lawrence C. Evans and Fraydoun Rezakhanlou. Previously, he was a CNRS researcher at the Institut de Mathématiques de Toulouse before joining Université Paris Cité. His habilitation thesis (2019) focuses on optimal transport applications in analysis and probability. Fathi's research centers on optimal transport theory, particularly its applications to analysis, probability, and statistical physics. Key topics include interacting particle systems, functional inequalities (e.g., Poincaré, log-Sobolev), high-dimensional phenomena, Ricci curvature in discrete/continuous spaces, Stein's method, concentration of measure, and numerical methods for stochastic dynamics. His work is supported by the ANR project 'Conviviality.' He has delivered courses on functional analysis at ENS and participated in summer schools, including an MSRI course on functional inequalities and localization techniques. His teaching materials include lecture notes on optimal transport and stochastic processes. His contributions have been recognized through awards such as the IUF membership. Notable research collaborations include work with Thomas Courtade, Matthias Erbar, and Gabriel Stoltz on topics ranging from stability estimates of inequalities to hypocoercivity and numerical analysis of stochastic systems.

Jia-Jie ZhuView profile

Researcher

• Machine Learning
• Optimization
• Optimal Transport
• +5 more

Jia-Jie Zhu is a machine learner and applied mathematician currently serving as head of an independent research group at the Weierstrass Institute for Applied Analysis and Stochastics in Berlin, with an upcoming appointment as tenured associate professor at KTH Royal Institute of Technology in Stockholm. Previously, he conducted postdoctoral research in machine learning at the Max Planck Institute for Intelligent Systems in Tübingen, Germany, following doctoral studies in optimization and numerical analysis at the University of Florida. His research focuses on the mathematical foundations of machine learning and optimization, particularly at the intersection of computational algorithms, dynamical systems, and probability theory. Key areas include: Robust probabilistic machine learning algorithms Kernel methods for distribution manipulation Variational methods for optimization over probability distributions Gradient flows and optimal transport theory Wasserstein and Fisher-Rao geometry Applications in generative modeling and causal inference Dr. Zhu's recent work reveals deep connections between partial differential equations, kernel methods, and machine learning, resulting in theoretically grounded algorithms for handling distribution shifts. His publications demonstrate a consistent trajectory from foundational optimization theory to cutting-edge applications in robust learning and generative modeling, with increasing emphasis on the mathematical structures underlying modern ML systems. He has secured significant research funding, including a DFG Project on 'Optimal Transport and Measure Optimization Foundation for Robust and Causal Machine Learning' within the Priority Program 'Theoretical Foundations of Deep Learning' (SPP 2298), and actively organizes academic events such as the Workshop on Optimal Transport from Theory to Applications (OT-DOM) and upcoming sessions at ICSP 2025 and SwissMAP. As an educator, Dr. Zhu teaches nonparametric statistics at Humboldt University of Berlin and serves as area chair for major conferences including AISTATS 2025. He maintains an active research group with opportunities for master's students, PhD candidates, and postdoctoral researchers interested in the mathematical frontiers of machine learning.

• Machine Learning
• Domain Adaptation
• Transfer Learning
• +3 more

Dr. Sinno Jialin Pan is a leading researcher in machine learning and artificial intelligence at Nanyang Technological University, Singapore. His work focuses on domain adaptation, sentiment analysis, and efficient neural network optimization. Key research areas: Machine Learning, Domain Adaptation, Reinforcement Learning, Sentiment Analysis Recent publications demonstrate expertise in time-series classification (2025) using hierarchical domain adaptation, LLM efficiency (2025) through expert pruning, and graph generation (2024) via spectral diffusion. His work spans both theoretical advancements and practical applications in neural architecture optimization and adversarial learning. Scientific contributions include: 2025: Virtual-label hierarchical domain adaptation 2024: Spectral diffusion for graph generation 2024: Multilingual jailbreak analysis in LLMs Current trends show increasing focus on large language model optimization and robust neural architectures , with applications in fault diagnosis, recommender systems, and misinformation detection.

• Process Mining
• Business Process Management
• Stochastic Processes
• +4 more

Dr. Sander Leemans is a Professor at RWTH Aachen University leading the Business Process Management Foundations and Engineering research group. His work focuses on advancing process mining theory and practice with emphasis on stochastic modeling and conformance verification. Leemans' research centers on process mining, business process management, and stochastic process modeling. He investigates conformance checking techniques for probabilistic models, process discovery algorithms, and the integration of exogenous data into process analysis. His work bridges theoretical foundations with practical applications in healthcare, robotic process automation, and inter-organizational systems. Recent publications reveal a concentrated research trajectory in stochastic conformance checking, where Leemans develops methods for matching observed traces to stochastic process models using alignment techniques, entropy metrics, and partial-order reasoning. He also pioneers object-centric process mining frameworks and explores silent transitions in labeled Petri nets, significantly enhancing the precision and applicability of process mining in real-world scenarios. The Business Process Management Foundations and Engineering group under Leemans' leadership drives innovation in process mining through rigorous theoretical development and open-source tooling, maintaining RWTH Aachen's position at the forefront of business process intelligence research.