Lilian Welschinger

Lilian Welschinger

Incoming PhD student at EPFL (CVLab) · currently research at INRIA Willow, ENS — Paris & Lausanne.

About

I am broadly interested in how machines can learn to understand and predict the physical world. My background is in mathematics; my current work sits at the intersection of generative models, physical dynamics, and representation learning. I am an EDIC fellow at EPFL, where I will pursue this research in the CVLab.

Publications

CVPR 2026

Learning to Solve PDEs on Neural Shape Representations

Lilian Welschinger, Yilin Liu, Zican Wang, Niloy J. Mitra

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026.

Project teaser

Education

Sept 2026

PhD, Computer & Communication Sciences (EDIC) — EPFL

Starting September 2026. EDIC Fellowship. Expected to join the CVLab.

2024–25

MVA (Mathematics, Vision, Learning) — ENS Paris-Saclay

Research master's. GPA 4.0 / 4.0.

2022–25

MSc, Data Science, Optimisation & Probability — ENSTA Paris (IP Paris)

GPA 4.0 / 4.0, top 5% of cohort.

Contact

email