What do you think?
Rate this book
Vision: A Computational Investigation into the Human Representation and Processing of Visual Information
David Marr's posthumously published Vision (1982) influenced a generation of brain and cognitive scientists, inspiring many to enter the field. In Vision, Marr describes a general framework for understanding visual perception and touches on broader questions about how the brain and its functions can be studied and understood. Researchers from a range of brain and cognitive sciences have long valued Marr's creativity, intellectual power, and ability to integrate insights and data from neuroscience, psychology, and computation. This MIT Press edition makes Marr's influential work available to a new generation of students and scientists.
In Marr's framework, the process of vision constructs a set of representations, starting from a description of the input image and culminating with a description of three-dimensional objects in the surrounding environment. A central theme, and one that has had far-reaching influence in both neuroscience and cognitive science, is the notion of different levels of analysis -- in Marr's framework, the computational level, the algorithmic level, and the hardware implementation level.
Now, thirty years later, the main problems that occupied Marr remain fundamental open problems in the study of perception. Vision provides inspiration for the continuing efforts to integrate knowledge from cognition and computation to understand vision and the brain.
In Marr's framework, the process of vision constructs a set of representations, starting from a description of the input image and culminating with a description of three-dimensional objects in the surrounding environment. A central theme, and one that has had far-reaching influence in both neuroscience and cognitive science, is the notion of different levels of analysis -- in Marr's framework, the computational level, the algorithmic level, and the hardware implementation level.
Now, thirty years later, the main problems that occupied Marr remain fundamental open problems in the study of perception. Vision provides inspiration for the continuing efforts to integrate knowledge from cognition and computation to understand vision and the brain.
428 pages, Paperback
First published January 1, 1982
About the author
David Marr
6 books9 followersLibrarian Note: There is more than one author in the GoodReads database with this name. See this thread for more information.
David Courtnay Marr was a British neuroscientist and psychologist. Marr integrated results from psychology, artificial intelligence, and neurophysiology into new models of visual information processing. He is acknowledged as a founder of the discipline of Computational Neuroscience.
David Courtnay Marr was a British neuroscientist and psychologist. Marr integrated results from psychology, artificial intelligence, and neurophysiology into new models of visual information processing. He is acknowledged as a founder of the discipline of Computational Neuroscience.
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 7 of 7 reviews
November 15, 2012
David Marr proposes a complete framework of how the brain could process visual information from 2D image all the way to 3D geometries at the very end. Even coming up with a not-obviously-wrong hypothesis of this entire pipeline is no small feat and David almost makes it sound consistent and as if it could work if it was only implemented with a few details filled in here and there. I do wonder if he was slightly ahead of his time, I'm sure he would have loved to play around with Kinect RGBD videos :)
December 2, 2015
Begin with the epilogue. If any of its arguments interest you, go find their proofs in the preceding text.
October 23, 2015
David Marr's Vision is a book of Marr's research on the study of the visual sytem. It was not fun to read and not for general audiences. It is easy to understand the basic premise of the book, however, and I think it's the right way to go about studying vision.
Marr's perspective on the study of vision amounts to this. When we study any system in the brain, we have to study it at three different levels. There's the computational level, the algorithmic level, and the physiological level (or neurophysiological level, if you'd like). Here is how to think about these different levels. The computational level tries to answer the question, What does the system do? What problem is the system designed to solve? In the case of vision, the question might be, What reason do we have the visual system in general or some of the features of the visual system in particular?
The algorithmic or representational level deals with these kinds of questions. How does the system do what it does? What representations in particular or what algorithms does the system make use of to carry out its processing?
The physical or physiological level asks how the system physically realizes the system? With vision, the question might be what neural structures are used to make vision possible? According to Marr, any neglect of any of these areas will make it very difficult to make any progress in the study of any given system.
Marr argues in this book that the general way in which vision works is that features of the world are interpreted first in a rough sketch form, kind of like a rough pencil drawing on paper. When it wants to understand the information that is being input in more detail, the visual system might then create a two and a half dimensional sketch, where textures and shades, for example, are interpreted in order to understand something in the environment. If still more information about the environment is needed, the visual system will construct a three dimensional model.
If we accept Marr's framework for a research program and his three stages of vision, then it makes possible to work out the details to study vision further, more richly and with more scientific rigor. That said, the writing here is torturous for general audiences because Marr uses language related to the visual system that only people who are familiar with the visual system would understand. Unfortunately.
Marr's perspective on the study of vision amounts to this. When we study any system in the brain, we have to study it at three different levels. There's the computational level, the algorithmic level, and the physiological level (or neurophysiological level, if you'd like). Here is how to think about these different levels. The computational level tries to answer the question, What does the system do? What problem is the system designed to solve? In the case of vision, the question might be, What reason do we have the visual system in general or some of the features of the visual system in particular?
The algorithmic or representational level deals with these kinds of questions. How does the system do what it does? What representations in particular or what algorithms does the system make use of to carry out its processing?
The physical or physiological level asks how the system physically realizes the system? With vision, the question might be what neural structures are used to make vision possible? According to Marr, any neglect of any of these areas will make it very difficult to make any progress in the study of any given system.
Marr argues in this book that the general way in which vision works is that features of the world are interpreted first in a rough sketch form, kind of like a rough pencil drawing on paper. When it wants to understand the information that is being input in more detail, the visual system might then create a two and a half dimensional sketch, where textures and shades, for example, are interpreted in order to understand something in the environment. If still more information about the environment is needed, the visual system will construct a three dimensional model.
If we accept Marr's framework for a research program and his three stages of vision, then it makes possible to work out the details to study vision further, more richly and with more scientific rigor. That said, the writing here is torturous for general audiences because Marr uses language related to the visual system that only people who are familiar with the visual system would understand. Unfortunately.
May 3, 2015
Difficult book to rate. I feel that it is a very personal and interesting viewpoint on the the perception of images, but I cannot shake the idea that it is also quite a bit outdated. It has a bit of an old AI feel where the focus was on algorithms rather than probability and networks. Contained some interesting concepts (for example, how we perceive depth and texture), but seemed to be both too technical and at the seem time too superficial to me...
December 20, 2021
I was not disappointed by the book in itself but was expecting better quality images and diagrams - I find it unbelievable the MIT Press sells a black-and-white, soft-cover book for 45$US. The first two chapters, including the introduction of the famous three-level framework were in my opinion the best part.
September 29, 2021
Seminal work for my master’s thesis.
April 29, 2024
I have a PhD in this topic and ngl the book if for sure informative but isnt going to appeal to those who already know their stuff...I kind of wish he sprinkled in a bit of personality ngl...
Displaying 1 - 7 of 7 reviews