Oriane Siméoni to present DINOv3 at CVPR 2026, arguing the case for scaling self-supervised learning in computer vision · Digg