Transformer-in-Vision

Why we included this project

Transformer-in-Vision is a curated reading list for anyone trying to keep up with how transformer architectures spread from NLP into computer vision and beyond. The entries span CLIP and DALL·E 2, video generation, 3D synthesis, and large multimodal models, and each one points back to the original paper or official code instead of a rehosted summary. That design makes it handy for two different readers: newcomers can browse the list to get oriented, while researchers and engineers can quickly locate the canonical source behind a method they want to explore. It is more of a map of the field than a tutorial, so it works best for people who know the basics and want to decide which directions deserve a closer look.