Lucas Degeorge
Scholar

Lucas Degeorge

Google Scholar ID: wSx6W9cAAAAJ
Ecole polytechnique
Computer visiondiffusion modelsconditional generative models
Homepage↗Google Scholar↗
Citations & Impact
All-time
Citations
1
 
H-index
1
 
i10-index
0
 
Publications
3
 
Co-authors
0
 
Contact
Emaillucas.degeorge@polytechnique.eduTwitterOpen ↗GitHubOpen ↗LinkedInOpen ↗
Publications
5 items
Dino Forcing Flow Models: Do not denoise what you can predict
2026
Cited
0
Balancing Frequencies and Pixels in Flow Matching
2026
Cited
0
PoM: A Linear-Time Replacement for Attention with the Polynomial Mixer
2026
Cited
0
MIRO: MultI-Reward cOnditioned pretraining improves T2I quality and efficiency
2025
Cited
0
One-step Diffusion Models with Bregman Density Ratio Matching
2025
Cited
0
Resume
Academic Achievements
  • Publications:
  • - MIRO: MultI-Reward cOnditioned pretraining improves T2I quality and efficiency
  • - How far can we go with ImageNet for text-to-image generation?
  • - One-step Diffusion Models with Bregman Density Ratio Matching
  • Involved in multiple projects and published works on arXiv.
Research Experience
  • Research at VISTA (Ecole polytechnique) and IMAGINE (Ecole des Ponts) labs; attended ECCV 2024 in Milano.
Education
  • PhD student in Computer Vision, Ecole polytechnique & Ecole des Ponts, supervised by Vicky Kalogeiton and David Picard.
Background
  • Main research interest: multi-modal generative models. Currently working on text-to-image generation.
Miscellany
  • Attended EurIPS 2025 to present work at the Workshop on Principles of Generative Modeling (PriGM).