Ivan De Boi - FWO postdoctoral fellow

About me

My research lies at the intersection of geometric machine learning, uncertainty quantification, and optimization under uncertainty. I develop probabilistic models and optimization methods that explicitly exploit the geometry of structured state spaces, including projective and manifold-valued representations. Rather than forcing complex systems into conventional Euclidean formulations, my work seeks to leverage their underlying geometric structure to improve learning, inference, calibration, and decision-making.

A central theme throughout my research is the development of geometry-aware probabilistic methods capable of quantifying uncertainty while respecting physical, geometric, or mathematical constraints. My methodological contributions span Gaussian processes, Bayesian optimization, geometric inverse problems, manifold-based representations, and uncertainty-aware machine learning.

These methods have been applied across a diverse range of domains, including computer vision, medical imaging, rehabilitation sciences, nuclear physics, and predictive maintenance. By bridging foundational applied mathematics with modern machine learning, I aim to develop principled frameworks for discovering, calibrating, and controlling complex systems under uncertainty.

Research Vision

While the field of geometric machine learning has made significant strides in shifting away from purely Euclidean assumptions, integrating these geometries deeply into probabilistic frameworks remains a critical frontier. Numerous scientific and engineering problems naturally live on structured spaces such as manifolds, projective spaces, geometric transformations, constrained state spaces, and subspaces, requiring methods that respect these intrinsic properties.

My long-term research objective is to establish a unified framework for geometry-aware probabilistic machine learning and optimization under uncertainty. This includes the development of Gaussian-process models, Bayesian optimization algorithms, and uncertainty quantification techniques that explicitly exploit geometric structure. Such methods offer new opportunities for formulating Bayesian inverse problems, improving sample efficiency, incorporating prior knowledge, and obtaining interpretable uncertainty estimates.

Ultimately, I seek to bridge geometry, probability, and optimization, enabling the n generation of uncertainty-aware learning systems for contemporary scientific challenges and industrial applications.

Research Interests

• Bayesian Optimization
• Uncertainty Quantification
• Geometric Machine Learning
• Gaussian Processes
• Bayesian Inverse Problems
• Projective and Manifold Geometry

Acquired Grants

Total: € 2,421,302.76 (excluding funding still under submission)

External (FWO, VLAIO, SCK CEN) € 1,928,416.21
Internal (IOF, BOF, OJO, PWO, AUHA, Innovation Fund) € 492,886.55

• Project Title: Predictive Maintenance for Non-Road Mobile Machinery
  Timeframe: September 2026 - August 2028
  Funding Source: VLAIO TETRA
  Funding Amount: € 480,000
  Role: supervisor, grant application 50/50 with KdG University College
  Status: Approved

• Project Title: Bayesian optimization algorithms for ISOL ion source optimal tunning and characterization
  Timeframe: September 2026 - September 2030
  Funding Source: SCK CEN
  Funding Amount: € 276,870
  Role: supervisor, grant application 25% me, 75% SCK CEN
  Status: Approved

• Project Title: Symposium on How to Exploit Uncertainty in Research II
  Timeframe: applied November 2025, symposium will be held on 3/9/26
  Funding Source: OJO Call UAntwerp
  Funding Amount: € 4,000
  Role: supervisor and grant holder
  Status: Approved

• Project Title: PreMaNRM2 – Predictive Maintenance for Non-Road Mobile Machinery
  Timeframe: applied June 2025, project November 2025 - October 2026
  Funding Source: AUHA Zaaifonds voor onderzoekssamenwerking
  Funding Amount: € 25,000
  Role: supervisor and grant holder
  Status: Approved, ongoing

• Project Title: Symposium on How to Exploit Uncertainty in Research
  Timeframe: applied November 2024, symposium was be held on 3/9/2025
  Funding Source: OJO Call UAntwerp
  Funding Amount: € 4,000
  Role: supervisor and grant holder
  Status: Completed

• Project Title: Implicit Camera Calibration
  Timeframe: applied November 2024, project April 2025 – October 2025
  Funding Source: BOF KP UAntwerp
  Funding Amount: € 10,000
  Role: supervisor and grant holder
  Status: Completed

• Project Title: Wireless capsule endoscopy based on Gaussian process latent variable models,
  Timeframe: October 2024 – September 2027
  Funding Source: FWO Postdoctoral Fellowship
  Funding Amount: € 380,000
  Role: grant holder, fellow
  Status: Approved, ongoing

• Project Title: Virtual Reality application for Treatment of Spatial Neglect
  Timeframe: September 2020 – August 20
  Funding Source: IOF POC UAntwerp
  Funding Amount: € 86,034.55
  Role: 50/50 effort with MOVANT University of Antwerp
  Status: Completed

• Project Title: AI in XR, toolkit for AI-development in XR-applications
  Timeframe: September 2019 – August 2020
  Funding Source: PWO Karel de Grote University College
  Funding Amount: € 138,223
  Role: supervisor and grant holder
  Status: Completed

• Project Title: ImmersiMed
  Timeframe: September 2019 – August 2020
  Funding Source: Innovation Fund Karel de Grote University College
  Funding Amount: € 37,860
  Role: supervisor and grant holder
  Status: Completed

• Project Title: Education 4.0
  Timeframe: September 2018 – August 20,
  Funding Source: VLAIO Living Lab (Proeftuinen)
  Funding Amount: € 343,968.00
  Role: supervisor for our part, grant application 25% me, 25% Catalisti, 25% IMEC, 25% ACTA
  Status: Completed

• Project Title: IoTTo, IoT-platform for real time vehicle data
  Timeframe: September 2019 – August 2020
  Funding Source: PWO Karel de Grote University College
  Funding Amount: € 191,765
  Role: grant application
  Status: Completed

• Project Title: ELGAS, Effecten van Luchtkwaliteit op de Gezondheid in Accommodaties van Schepen
  Timeframe: September 2019 – August 2020
  Funding Source: VLAIO TETRA
  Funding Amount: € 197,154.75
  Role: grant application 25% me, 50% Antwerp Maritime Academy (Hogere Zeevaartschool), 25% VITO
  Status: Completed

• Project Title: Collaborative XR Lab
  Timeframe: September 2018 – August 2020
  Funding Source: VLAIO, Steun voor onderzoeksinfrastructuur bij hogescholen
  Funding Amount: € 250,423.46
  Role: supervisor, grant application 50/50 effort with coworker
  Status: Completed

Scientific Output and Impact

Patent

Patent: COMPUTER IMPLEMENTED METHOD AND SYSTEM FOR MAPPING SPATIAL ATTENTION Publication number: WO2022/122834

Patentscope

YouTube video

Invited Talks

• 27/3/26, Gaussian Processes, Eindhoven University of Technology (NL), 16 attendees
• 18/12/25, Introduction to Gaussian Processes, University of Namur, 30 attendees
• 08/09/25, Constraints on GP Predictions, Gaussian Process Summer School 2025, University of Manchester, 40 attendees
• 25/08/25, Bayesian Optimisation Summer School 2025, Hasselt University, 30 attendees
• 20/01/25, How to do more with uncertainty in AI, Applied Computer Science Conference Days, Karel de Grote University College, >100 attendees
• 4/12/24, How to Exploit Uncertainty in AI for XR development, Multimedia and Creative Technologies, Karel de Grote University College, >100 attendees
• 23/01/24, How to do more with uncertainty in AI, Applied Computer Science Conference Days, Karel de Grote University College, 100 attendees
• 24/5/23, On straight lines and machine learning, Systems and Robotics of the university of Coimbra (Portugal), 15 attendees
• 1/2/23, Straight lines in robotics, Machine Learning Research Group of Oxford (UK), 10 attendees
• 16/11/18, Unlock the power of AR, Applied Computer Science Conference Days, Karel de Grote University College, 100 attendees
• 16/11/17, Jumpstart into Hololens development with Unity, Applied Computer Science Conference Days, Karel de Grote University College, 100 attendees

Conference Talks

• 30/6/26, Beyond Euclidean: A Conceptual Review of Distance Metrics and Dissimilarity Measures in Contemporary Machine Learning, 20th Annual International Conference on Mathematics: Teaching, Theory & Applications, Athens (Greece)
• 15/07/25, The creation of ruled surfaces in a hall of mirrors, Bridges 2025 : Mathematics and the Arts, 40 attendees
• 22/10/24, Gauss is GOAT @Mathfest24, 100 attendees
• 27/11/23, How to turn your camera into a perfect pinhole model, 26th Iberoamerican Congress on Pattern Recognition at Coimbra (Portugal), 80 attendees
• 5/10/23, How to get more out of endoscopic images, Cancer Research Day, University of Antwerp, 100 attendees
• 7/4/22, How to improve the accuracy of laser-based systems using straight lines, Metromeet conference (Spain), 50 attendees
• 7/9/20, A taxonomy of contemporary artificial intelligence techniques used in virtual reality training and simulations, SciFi-It conference, 20 attendees
• 12/12/19, Gaussian Processes – Literature Study 2019, lunch seminar hosted by InViLab University of Antwerp, 32 attendees
• 18/4/18, ImmersiMed: a future-proof immersive platform for medical training at the SciFi-It conference 20 attendees

Science Communication

• 4/03/24, AI skeleton tracking fitness mirror, Flanders Technology and Innovation closing event, >1000 attendees
• 24/11/19, Dag van de wetenschap, VR demonstration ImmersiMed, an application to train medical personnel, >1000 attendees
• 29/9-30/9 18, ImmersiMed, Tech and Innovation Festival SuperNova, Flanders DC, >2000 attendees
• 25/11/18, Dag van de wetenschap, AR demonstration on Smart Vending Machines, >1000 attendees

Network on Gaussian Processes

In December 2019 we launched an interdisciplinary interest group regarding the possibilities and the challenges of Gaussian Processes. The objective was (and still is) the organic expansion of a platform for discussions and sharing of ideas on theoretical contributions and practical applications, creating a fertile soil for future cooperation and project proposals. We have had the pleasure to welcome various speakers such as professor Søren Hauberg (Technical University of Denmark), professor Stephen Roberts (Oxford University), professor Carl Henrik Ek (Cambridge University) and many more. Find all of them here. Do not hesitate to contact me if you want to get involved.

Academic Service and Memberships

• 2026 July - Session Chair, 20st Annual International Conference on Mathematics: Teaching, Theory & Applications, Athens, Greece, https://www.atiner.gr/2026mat-pro
• 2025 - ongoing, Member of COST Action Extended Reality Neurorehabilitation of Spatial Neglect and Related Disorders After Brain Injury (neuroXRehab), WG2 Assessment of spatial neglect and related disorders and WG3 Treatment of spatial neglect and related disorders, www.cost.eu/actions/CA22137
• 2025 - ongoing, Member of Management Committee COST Action Adaptive and acquired resistance in gastrointestinal cancers-contemporary and emerging resolutions (AdResCanCER), WG5 Integrated data analyses and machine learning, https://www.cost.eu/actions/CA24162/
• 2024 - ongoing, BAP representative Faculty Educational Council (OWC)
• 2024 - ongoing, Vlaamse Interuniversitaire Raad (VLIR), postdoc representative for University of Antwerp
• 2024 - ongoing, Member of COST Action Randomised Optimisation Algorithms Research Network, WG4 Optimisation under uncertainty, www.cost.eu/actions/CA22137 https://www.roar-net.eu/
• 2022-2024 PhD student representative in both board and council of Antwerp Doctoral School
• 2022-2024 PhD student representative in Faculty Doctoral Committee (FDOC)
• 2022-2024 PhD student representative in Faculty Research Council (OZC)

Awards and Honours

• Our paper “Probabilistic Camera Distortion Correction Using Deep Gaussian Processes” was selected for the cover story of JImaging, Volume 12, Issue 7 (July 2026).
• I won the audience award for my pitch on Gauss as GOAT @MathFest24.
• Best paper award for my paper “IMMERSIMED: A Future-Proof Immersive Platform for Medical Training” at the SCIFI-IT’2018 Conference (April 18-20 2018).
• Our paper “Dynamic Line Scan Thermography Parameter Design via Gaussian Process Emulation” was selected for the cover story of Algorithms, Volume 15, Issue 4 (April 2022)
• In 2019, I received the audience best project award (publieksprijs) “Prijs Frank Swaelen”, at the Karel de Grote Hogeschool congress for my Innovation Project “ImmersiMed - training hospital procedures in VR”.

Reviewing Activities

• IEEE Transactions on Instrumentation and Measurement
• Elsevier Measurement
• Springer Autonomous Intelligent Systems and Springer Nonlinear Dynamics
• Springer Nonlinear Dynamics
• International Conference on Control, Mechatronics and Automation (editions 2024, 2025, 2026), ICCMA
• BMVC 2025, The 36th British Machine Vision Conference
• I also review for VLAIO (afdeling Innovatiesteun van het Agentschap Innoveren & Ondernemen) as an expert on AI and AR/VR.

Publications

2026

A1 Journal

Probabilistic Camera Distortion Correction Using Deep Gaussian Processes, Ivan De Boi, Rhys G. Evans, Stuti Pathak, Thomas De Kerf, Marnix Van Soom, Sam Van der Jeught, Helder Araújo, and Rudi Penne. 2026. Journal of Imaging 12, no. 7: 296. https://doi.org/10.3390/jimaging12070296

A Unified Complex-Fresnel Model for Physically Based Long-Wave Infrared Imaging and Simulation, ter Heerdt, Peter, William Keustermans, Ivan De Boi, and Steve Vanlanduit. 2026. Journal of Imaging 12, no. 1: 33. https://doi.org/10.1016/j.nimb.2025.165859

2025

A1 Journal

Isotope Separator On-Line system tuning: Bayesian optimization applied to the transport beamline case, Santiago Ramos Garces, Line Le, Mia Au, Alexander Schmidt, João Pedro Ramos, Marc Dierckx, Dinko Atanasov, Ivan De Boi, Sebastian Rothe, Lucia Popescu, Stijn Derammelaere. 2025. Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms. 2025/11/1. Volume 568. https://doi.org/10.1016/j.nimb.2025.165859

Use of Immersive Virtual Reality to Explore Visual Search Behaviour in Individuals with Visuospatial Neglect after Stroke, Elissa Embrechts, Ivan De Boi, Quirine Schatteman, Tanja C. W. Nijboer, Steven Truijen, and Wim Saeys. 2025, Neuropsychological Rehabilitation, June, 1–28. https://doi:10.1080/09602011.2025.2511193.

On the Measurement of Laser Lines in 3D Space with Uncertainty Estimation, Ivan De Boi, Nasser Ghaderi, Steve Vanlanduit, Rudi Penne, Sensors 2025, 25, 298. https://doi.org/10.3390/s25020298

P1 Conference

The Creation of Ruled Surfaces in a Hall of Mirrors, Ivan De Boi, Rudi Penne, Proceedings of Bridges 2025 : Mathematics and the Arts - ISSN 1099-6702 - Tessellations Publishing, 2025, p. 483-486, https://archive.bridgesmathart.org/2025/bridges2025-483.html#gsc.tab=0.

Safe Bayesian optimization with simulation-informed Gaussian process for the constraints, Santiago Ramos Garces, Ivan De Boi, João Pedro Ramos, Marc Dierckx, Lucia Popescu, Stijn Derammelaere, ICAAI ‘24: Proceedings of the 2024 8th International Conference on Advances in Artificial Intelligence, Pages 66 - 72, https://doi.org/10.1145/3704137.3704157

2024

PhD thesis

Gaussian processes for 3D measurements, De Boi Ivan, Dissertation. University of Antwerp, 2024, https://repository.uantwerpen.be/desktop/irua

A1 Journal

Geometric optimization through CAD-based bayesian optimization with unknown constraint, Abdelmajid Ben Yahya, Robbe De Laet, Santiago Ramos Garces, Nick Van Oosterwyck, Ivan De Boi, Annie A.M. Cuyt, Stijn Derammelaere, 2024 12th International Conference on Control, Mechatronics and Automation (ICCMA), 11-13 November, 2024, London, United Kingdom, p. 394-402, https://doi.org/10.1109/ICCMA63715.2024.10843905

Efficient Tuning of an Isotope Separation Online System Through Safe Bayesian Optimization with Simulation-Informed Gaussian Process for the Constraints, Garces, Santiago Ramos, Ivan De Boi, João Pedro Ramos, Marc Dierckx, Lucia Popescu, and Stijn Derammelaere, 2024, Mathematics 12, no. 23: 3696. https://doi.org/10.3390/math12233696

Mechanism design optimization through CAD-based Bayesian optimization and quantified constraints, Ben Yahya Abdelmajid, Ramos Garces Santiago, Van Oosterwyck Nick, De Boi Ivan, Cuyt Annie, Derammelaere Stijn, Discov Mechanical Engineering 3, 21 (2024). https://doi.org/10.1007/s44245-024-00054-7

Assessment and treatment of visuospatial neglect using active learning with Gaussian processes regression, De Boi Ivan, Embrechts Elissa, Schatteman Quirine, Penne Rudi, Truijen Steven, Saeys Wim, Artificial intelligence in medicine - ISSN 0933-3657 - 149(2024), https://doi.org/10.1016/J.ARTMED.2024.102770

P1 Conference

Simulation-informed Gaussian Processes for Accelerated Bayesian Optimisation, Santiago Ramos Garces, Ivan De Boi, Rudi Penne, João Pedro Ramos, Marc Dierckx, and Stijn Derammelaere, Intelligent Management of Data and Information in Decision Making. August 2024, 219-226 https://doi.org/10.1142/9789811294631_0028

2023

A1 Journal

A phase correlation based peak detection method for accurate shape from focus measurements, Gladines Jona, Sels Seppe, De Boi Ivan, Vanlanduit Steve, Measurement / International Measurement Confederation - ISSN 1873-412X - 213(2023), https://doi.org/10.1016/J.MEASUREMENT.2023.112726

Enhanced checkerboard detection using Gaussian processes, Hillen Michaël, De Boi Ivan, De Kerf Thomas, Sels Seppe, Cardenas Edgar, Gladines Jona, Steenackers Gunther, Penne Rudi, Vanlanduit Steve, Mathematics - ISSN 2227-7390 - 11:22(2023), https://doi.org/10.3390/MATH11224568

On the angular control of rotating lasers by means of line calculus on hyperboloids, Penne Rudi, De Boi Ivan, Vanlanduit Steve, Sensors - ISSN 1424-8220 - 23:13(2023), https://doi.org/10.3390/S23136126

Surface approximation by means of Gaussian process latent variable models and line element geometry, De Boi Ivan, Ek Carl Henrik, Penne Rudi, Mathematics - ISSN 2227-7390 - 11:2(2023), https://doi.org/10.3390/MATH11020380

P1 Conference

How to turn your camera into a perfect pinhole model, De Boi Ivan, Pathak Stuti, Oliveira Marina, Penne Rudi, Lecture notes in computer science - ISSN 0302-9743 - Springer, (2024), p. 90-107, https://doi.org/10.1007/978-3-031-49018-7_7

2022

A1 Journal

Dynamic line scan thermography parameter design via Gaussian process emulation, Verspeek Simon, De Boi Ivan, Maldague Xavier, Penne Rudi, Steenackers Gunther, Algorithms - ISSN 1999-4893 - 15:4(2022), https://doi.org/10.3390/A15040102

Input and output manifold constrained Gaussian process regression for galvanometric setup calibration, De Boi Ivan, Sels Seppe, De Moor Olivier, Vanlanduit Steve, Penne Rudi, IEEE transactions on instrumentation and measurement / Institute of Electrical and Electronics Engineers - ISSN 1557-9662 - 71(2022), https://doi.org/10.1109/TIM.2022.3170968

Semi data-driven calibration of galvanometric setups using Gaussian processes, De Boi Ivan, Sels Seppe, Penne Rudi, IEEE transactions on instrumentation and measurement / Institute of Electrical and Electronics Engineers - ISSN 1557-9662 - 71(2022), https://doi.org/10.1109/TIM.2021.3128956

2020

A1 Journal

Feasibility of Kd-trees in Gaussian process regression to partition test points in high resolution input space, De Boi Ivan, Ribbens Bart, Jorissen Pieter, Penne Rudi, Algorithms - ISSN 1999-4893 - 13:12(2020), https://doi.org/10.3390/A13120327

P2 Conference

A taxonomy of contemporary artificial intelligence techniques used in virtual reality training and simulations, De Boi Ivan, Jorissen Pieter, International Science Fiction Prototyping Conference 2020(SciFi-It’2020), 23-25 March, 2020, Ghent, Belgium - ISBN 978-94-92859-10-5 - Ostend, EUROSIS-ETI, 2020, p. 41-48

2018

A1 Journal

ImmersiMed: Cross-platform simulation training, Jorissen Pieter, De Boi Ivan, http://www.digitmedicine.com/text.asp?2018/4/4/166/248975

PhD Students

• Noa Eeckhaut-Racon, “Bayesian optimization algorithms applied to ISOL ion source tuning and characterization”, Expected submission of thesis (quarter/year): 04/2030
• Matteo Impieri, “Design and calibration of AI-powered camera systems for armoured vehicles”, Expected submission of thesis (quarter/year): 04/2029
• Mengdi Lu, “The Assessment and Treatment of Visuospatial Neglect After Using A Self-developed Virtual Reality Application”, Expected submission of thesis (quarter/year): 04/2027
• Marnix Van Soom, “Bayesian Nonparametric Glottal Inverse Filtering”, Expected submission of thesis (quarter/year): 04/2026
• Santiago Ramos Garces, successfully defended “Enhanced performance of automatic tuning in isotope separation online systems through Bayesian optimization” on 1/12/2025

Work Experience

• Sep 2024 - ongoing, postdoctoral research fellow FWO
• May 2024 - Sep 2024, postdoctoral researcher University of Antwerp
• Feb 2021 - May 2024, PhD student University of Antwerp
• Feb 2009 - Jan 2021, Lecturer and Research Coordinator, Karel de Grote Hogeschool
• Feb 2017 - Jun 2017, Unity Developer, BUT (part time)
• Feb 2002 - Feb 2007, Application Engineer VRT
• Mar 2007 - Jan 2009, Project Engineer Test & Maintenance, Atlas Copco

Mock Publications

Do not take these seriously. Although, …

• Happy Edward Goodwin Day, aka the more sensible pi-day pdf
• On the Tardiness of my Coworkers and How To Exploit it pdf. Also published as a chapter in the book How to Prove Anything: 30 absurd research papers no one else was brave enough to publish. https://www.amazon.com/How-Prove-Anything-research-publish/dp/1806118939 Check out the Journal of Astrological Big Data Ecology https://jabde.com/