Séminaire IRA2 – Modelling the Spatio-temporal Variability in the Geometry and Structure of Tree-like Objects: From Statistical Analysis to Generative Models
Pr. Hamid LAGA, Murdoch University (Australie).
Abstract
Objects in nature evolve over time as the result of their interaction with the environment, natural growth, and disease progression. In this process, both their 3D geometry and structure deforms. Understanding and mathematically modelling such spatio-temporal deformations can benefit many applications, from health and biology to computer graphics and virtual worlds creation. In this talk, I will review our recent works that attempted to bridge the gap between traditional techniques for statistically modelling such spatio-temporal variabilities, using concepts from Riemannian shape spaces and metrics, and modern deep learning techniques that learn latent spaces.
- Date: 08/10/2026, 14h
- Lieu: IBISC, site Pelvoux, amphithéâtre Yasmina Bestaoui (Bx30), et à distance via Microsoft Teams
- Organisation: Hédi TABIA (PR Univ. Évry, IBISC équipe IRA2), Amine CHELLALI (MCF Univ. Évry, IBISC équipe IRA2)
