Stéphane Lathuilière

Research Scientist

Stéphane Lathuilière

I am a research scientist in the RobotLearn team at Inria Grenoble. My research interests include generative models for image and video generation (pre- and post-training), multimodal LLMs, and continual learning.

I have published papers in the most prestigious computer vision conferences (CVPR, ICCV, ECCV, NeurIPS) and top journals (T-PAMI). I have collaborated with numerous academic and industry leaders, including Snapchat, Meta, and Atos.


Last News


Publications

Full and updated list available on Google Scholar.


Community Service

I frequently contribute to the community as an Area Chair for major venues, including CVPR (2025-2026), ECCV (2024-2026), ICIAP 2025, and ACM Multimedia 2022. Additionally, I regularly serve as a reviewer for top-tier journals such as T-PAMI, IJCV, TIP, RA-L, and TMLR, as well as for international conferences including CVPR, ICCV, ICLR, ICML, NeurIPS, and ECCV. I served as an Associate Editor for IEEE Transactions on Multimedia (TMM) from 2023 to 2025.


Teaching

Since 2025, I have been involved in teaching activities at UGA and Ensimag-Grenoble INP, focusing on Deep Learning and Computer Vision (~40h/year). Previously, from 2019 to 2024, I was an Associate Professor at Télécom Paris, where I taught various programs in AI and Computer Vision. I remain occasionally involved in the AI Master's programs at both Télécom Paris and École Polytechnique.


Team


Contact

655 Av. de l'Europe, 38330 Montbonnot-Saint-Martin
Office H118