Cinema and Machine Vision unfolds the aesthetic, epistemic, and ideological dimensions of machine-seeing films and television using computers. With its critical-technical approach, this book presents to the reader key new problems that arise as AI becomes integral to visual culture. It theorises machine vision through a selection of aesthetics, film theory, and applied machine learning research, dispelling widely held assumptions about computer systems designed to watch and make images on our behalf.
At its heart, Cinema and Machine Vision is an invitation for film and media scholars to critically engage with AI at a technical level, a prompt for scientists and engineers working with images and cultural data to critically reflect on where their assumptions about vision come from, and a joint recognition of the fruitful problems of working together to understand the algorithmic governance of the visual.
Brings together Film Studies and Artificial Intelligence by exploring their mutual interest in the automation of vision through technology.
Arvustused
Cinema and machine vision rescaled the world, from cell to pixel, detail to data. Daniel Chavez Heras shows us the parallel and diverging ways they have become worldmaking. A wonderful and necessary book that starts the story of AI where it should be started: in much earlier technical imaging practices of cinema. -- Professor Jussi Parikka, Aarhus University, author of Operational Images
List of Illustrations
Acknowledgements
Introduction
Part One: Data-Images: Philosophy of Photography and Technologies of Vision
Chapter
1. Between Archive and Dataset
Chapter
2. Inductive Vision
Chapter
3. Machine Learning and the Philosophy of Photography
Part Two: Pixels in Motion: The Calculation of Cinematic Time
Chapter
4. Statistical Distance and Emotional Closeness in Film Style
Chapter
5. Computational Analysis of Continuity Editing
Chapter
6. Duration, Motion, and Pixels
Part Three: AI and Criticism: Aesthetics, Formats, and Interactions
Chapter
7. Algorithmic Films as Data Analysis
Chapter
8. Aesthetic Judgements and Meaningful Dissensus
Chapter
9. AI as Media
Conclusion: Machines Made of Images
References
Index
Daniel Chávez Heras is a Lecturer in Digital Culture and Creative Computing at Kings College London. He specialises on the computational production and analysis of visual culture combining critical frameworks in the history and theories of cinema, television, and photography, with advanced technical practice in creative and scientific computing, including applied machine learning technologies.Daniel has worked extensively in interdisciplinary design and creative industries, in Mexico and in the UK, with cultural institutions such as The British Council, and the BBC. He is a member of the Creative AI Lab, in partnership with the Serpentine Gallery, and part of the Computational Humanities Research Group at Kings College London.
