DINO

DINO (Self-Distillation with No Labels) is an artificial intelligence technique that enables computers to learn visual representations from images without needing human-provided labels. It works by having the model compare different versions of the same image—like different perspectives or augmentations—and learn to recognize features that stay consistent. This self-teaching approach helps the AI develop a nuanced understanding of images, making it useful for tasks like image classification or object detection. DINO leverages a process called self-distillation, where the model acts both as a student and a teacher, improving itself through internal comparison without external guidance.