This workshop introduces computer vision (and extent multimodal AI) through example and hands-on exercises, taught at an introductory level. It will focus largely on examples of collaborations between institutions and researchers in LAM and the Visual Geometry Group (VGG), a leading AI/ML research group at the University of Oxford. Participants will learn how to segment, classify, match and annotate various kinds of images and audivisual content: examples will be drawn from architecture, archaeology, book history, art history, film and photography. The workshop will demonstrate and teach the state of the art in Visual AI, including also critical and ethical issues such as bias and copyright, and the practicalities of collaborations. The session will be taught using free and open-source software: no coding experience is required, but participants will need a laptop and a Google Account to (optionally) access Colab notebooks.