05/10/2026
Dear friends, students and colleagues,
I would like to invite you to the first session of our Bucharest Computer Vision Reading Group this academic year, where I will present and we will discuss a very important, interesting and timely topic:
Models that learn a “semantic world model” using JEPA architectures, pioneered and developed by Yann LeCun’s research group.
I will focus on selected material from the following three papers:
* LeCun, Yann. “A path towards autonomous machine intelligence version 0.9. 2, 2022-06-27.” Open Review 62.1 (2022): 1–62.
* Assran, Mahmoud, et al. “Self-supervised learning from images with a joint-embedding predictive architecture.” 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2023.
* Assran, Mido, et al. “V-JEPA 2: Self-supervised video models enable understanding, prediction and planning.” arXiv preprint arXiv:2506.09985 (2025).
The papers are easy to find online, for example through Google Scholar.
Where: Room 705 (7th floor), PRECIS Building at Universitatea POLITEHNICA din București with access from Iuliu Maniu Boulevard.
When: Tuesday, October 6, at 10:00 AM.
We look forward to seeing you and hope that many of you can join us! 🙂
Have a wonderful day,
Marius
