FrescoDiffusion Best Paper
4K image-to-video with prior-regularized tiled diffusion.
Hugo Caselles-Dupré
Research Director at Obvious Research, in partnership with ISIR, Sorbonne University.
Co-founder of Obvious, a collective of artists and researchers creating art with machine learning.
My research focuses on generative AI for artistic creation: images, video and sound. Since 2023, I have led our research laboratory, supported by an ANR grant for new creative algorithms.
New Research
JoLT Oral · ECCV 2026
Joint latent trajectories for high-resolution generation.
Obvious
We develop our own AI algorithms and use them to make artworks, from the Belamy portraits to Japanese prints, masks, immersive installations and images of mental imagination.
Explore our work → Obvious website ↗
Topics
Generative AI
Image, video and sound generation.
Mind-to-Image
Brain activity & visual imagination.
Creative synthesis
Algorithms for our artworks at Obvious.
Learning agents
Perception, interaction and learning.
