The Image-Theories behind Computer Vision

The Image-Theories behind Computer Vision,
Leonardo Impett, Assistant Professor at the Department of Computer Science, University of Durham

You can see this video (enhanced version) in fullscreen here

Abstract
Every computer vision algorithm is a set of instructions for how to decode images - in other words, a programmatic way of seeing. Every such algorithm therefore, explicitly or otherwise, embodies a particular understanding of theories, philosophies and ideologies of vision. Since the 1980s, computer vision scientists have chiefly referred to David Marr’s Vision (1982), itself an account of human visual neuroscience from the perspective of information theory. Other perspectives were toyed with by the scientific community, such as James Gibson’s Ecological Approach to Visual Perception (1979); the choice is axiomatic, and its implications for computer vision are totalizing, as they have to do not only with technical solutions but with the research problems being posited in the first place.

This paper has two ambitions. Firstly, to uncover the tacit visual theories implicit in computer vision systems that might be used by humanities researchers. And secondly, that the very relationship between a theory of vision and its algorithmic implementation might be the start of a new intersection between computer vision and the history of visual culture: where we prototype computer vision systems based on historical theories of vision. These serve not as tools to aid the digestion of masses of visual material, but rather as algorithmic thought-experiments for exploring the intellectual history of vision.

Biography
Leonardo Impett is Assistant Professor of Computer Science at Durham University in the UK. He has a background in information engineering and machine learning, having worked/studied with the Cambridge Machine Learning Lab, the Cambridge Rainbow Group, and Microsoft Research Cairo. His Ph.D., with Sabine Süsstrunk and Franco Moretti at EPFL, was on the use of computer vision for the "distant reading" of the history of art. In 2018 he was a DH fellow at Villa I Tatti - the Harvard University Center for Italian Renaissance Studies. From 2018-2020 he was Scientific Assistant, then Scientist, at the Bibliotheca Hertziana - Max Planck Institute for Art History in Rome. Alongside his research in digital art history, he frequently works with machine learning in cultural institutions, like the Liverpool Biennial, the Royal Opera House, and the Whitney Museum of American Art.