About
DINO is a family of self-supervised vision models developed by Meta AI (FAIR). The name stands for "self-DIstillation with NO labels". Successive generations — DINO (2021), DINOv2 (2023) and DINOv3 (2025) — train vision backbones (Vision Transformer, ConvNeXt) without labels, producing general-purpose, high-quality image features useful for classification, detection, segmentation and depth estimation without fine-tuning.
