Glossary: Computer Vision Beyond the Basics

Transformer

A network design built from attention layers; it is the basis of today's language models, ViT, CLIP and DETR.