Computer Vision as Culture: Experiments with Algorithmic Seeing
by Gabriel Pereira and Bruno Moreschi
- MEDIA : ART : WRITE : NOW
- Published: forthcoming
- ISBN: 978-1-78542-173-0
- PDF ISBN: 978-1-78542-172-3
Computer Vision — the AI algorithms that interpret and analyse images and video — has become a part of everyday life, being present in everything from facial recognition and automated content moderation to autonomous vehicles. These systems are often explained in technical terms, through notions of accuracy, efficiency and scale. This book proposes to understand Computer Vision as culture instead. Attending to the cultural values and practices that shape Computer Vision reveals the potentials and limitations of these algorithms, especially in this period of industry-led techno-utopianism, as manifested in the beliefs about what these systems can and should do.
Tracing how Computer Vision algorithms are shaped by particular values, histories and ways of seeing, the book analyses ‘hegemonic CV’: the dominant paradigm of algorithmic seeing which flattens the visual world to serve as a tool of surveillance and commerce, while presenting itself as neutral. This paradigm takes shape in the bounding boxes that constrain how images are read, the hidden labour that makes machines see and the vast scale at which they are trained. Drawing on the concepts and tools of art and visual culture, the book asks what Computer Vision makes visible, what it simplifies and what it leaves out. The authors argue that cultural and artistic practices are vital sites for imagining otherwise — and for developing counter-hegemonic forms of computational seeing. Cultural spaces such as museums and galleries serve for them as laboratories for staging interventions that reimagine how algorithmic seeing might work.
Author Bios
Bruno Moreschi is an artistic researcher and Associate Professor in New Media in the Department of Art and Media at Aalto University’s School of Arts, Design and Architecture, Finland. His investigations are related to the deconstruction of systems and the decoding of procedures and social practices in art spaces, museum collections, visual culture and technology more broadly. He has been a researcher at numerous universities and scholarly institutes: LIAS – Leuphana University Institute for Advanced Studies, Collegium Helveticum (ETH Zurich), Hanse-Wissenschaftskolleg, University of São Paulo, C4AI Inova USP and University of Cambridge. His projects have been recognised by grants and exhibitions from many institutions: ZKM, Volkswagen Foundation, Van Abbemuseum, 33rd Sao Paulo Art Biennial, FAPESP, SESC, Rumos Itaú Cultural, Funarte, CAD+SR and the Bauhaus Fellowship.
Gabriel Pereira is Assistant Professor in AI & Digital Culture at the University of Amsterdam in the Netherlands, based in the Media Studies department and the Institute for Logic, Language and Computation (ILLC). His research focuses on critical studies of data, algorithms and digital infrastructures, particularly those of computer vision. His research methods are interdisciplinary and collaborative, and include both qualitative research and practice-based inquiry (e.g. artistic interventions and critical making). His collaborative projects have been exhibited at the 33rd Sao Paulo Art Biennial, the Van Abbemuseum, IDFA DocLab and Itaú Cultural.