Tag: #DINO
Kubernetes, DevOps and MLOps, written from things we ran and measured ourselves · 1 posts
Self-Supervised Learning 1: From SimCLR and MoCo to DINO and DINOv2 — How to Learn Features Without Labels and Prevent "Collapse"
This post explains self-supervised learning, which learns useful features from a pile of unlabeled images, in the order of SimCLR and MoCo (negative examples and pushing apart), BYOL (no negative examples, predictor, and
2026-10-17 · 16 min read #비전 모델#자기지도 학습#SimCLR#MoCo#BYOL