Simon Coessens
I work on computer vision and generative models, with a recent focus on diffusion-based image generation.
Education
- MSc Computer Science, CentraleSupélec, 2023–2025 — Erasmus Mundus scholarship, mention Très bien, with mobility across:
- École Polytechnique de Bruxelles (ULB)
- Universitat Politècnica de Catalunya (UPC)
Thesis: 360° Panorama Generation from Arbitrarily Posed Input Images using Multi-View Diffusion. Supervisor: Akash Malhotra. - BSc Mathematics, KU Leuven, 2020–2025 — cum laudeTaken alongside Informatics degree; averaged ~85 ECTS credits per year. Thesis: Forecasting Stationary Processes. Supervisor: Dr. Sergey Berezin.
- BSc Informatics, KU Leuven, 2019–2023 — cum laudeThesis: Efficiency Analysis of Compressed Sensing in a Boolean Setting. Supervisor: Prof. Daan Huybrechs.
Publications
- GaussianDiffusion: Learning Image Generation Process in Gaussian Representation SpaceSimon Coessens, A. Samal, A. Malhotra, N. Bennacer Seghouani. 2025.
- MultiViewPano: A Generalist Approach to 360-degree Panorama GenerationSimon Coessens, A. Malhotra, N. Bennacer Seghouani. Structural Priors for Vision (SP4V) Workshop, ICCV 2025 — Spotlight. [OpenReview]
Links
