Lunchtime Lectures mit Dr. Giles Bergel (Visual Geometry Group, Oxford)

Computer vision and the printed book how to detect, match, compare and classify book parts

Digitisation and the development of software tools for the study digital images,

most recently those employing various forms of machine learning, have greatly

advanced the study of early printed books. Some new techniques address longstanding challenges, such as the detection of printing variants; others open up new areas of research, including large-scale classification of the printed page and its component parts over time. This talk will present ways that the speaker and colleagues have used these new tools, demonstrating the state of the art in the artificial understanding of the printed page.