An overview of Transformers and how to use them for vision tasks

Presentation

Code

Introduced in 2017, Transformers have come to dominate the Natural Language Processing field in just a few years. While the architecture was first applied to language tasks, in this workshop we will explore how this very general neural network architecture can be applied to arbitrary tasks and in particular how they can be used for computer vision. We will explore recent trends in vision Transformers with examples from medical imaging.