Marlene Behrmann
- Published in print:
- 2006
- Published Online:
- March 2012
- ISBN:
- 9780195313659
- eISBN:
- 9780199848058
- Item type:
- chapter
- Publisher:
- Oxford University Press
- DOI:
- 10.1093/acprof:oso/9780195313659.003.0012
- Subject:
- Psychology, Cognitive Psychology
This chapter takes a neuropsychological perspective on questions concerning analytic and holistic processing. It examines the behavior of seven brain-damaged patients who have problems with ...
More
This chapter takes a neuropsychological perspective on questions concerning analytic and holistic processing. It examines the behavior of seven brain-damaged patients who have problems with perceptual organization. These “integrative agnostic” patients seem to be disproportionately impaired on tasks tapping holistic configural processes compared to part-based processes. The first section of this chapter outlines three main empirical issues falling under the domain of perceptual organization: figure-ground organization, visual interpolation, and grouping. The second section contains a description of the patients. The third section examines the nature of the impairment in perceptual organization, in relation to figure-ground organization, visual interpolation, and grouping.Less
This chapter takes a neuropsychological perspective on questions concerning analytic and holistic processing. It examines the behavior of seven brain-damaged patients who have problems with perceptual organization. These “integrative agnostic” patients seem to be disproportionately impaired on tasks tapping holistic configural processes compared to part-based processes. The first section of this chapter outlines three main empirical issues falling under the domain of perceptual organization: figure-ground organization, visual interpolation, and grouping. The second section contains a description of the patients. The third section examines the nature of the impairment in perceptual organization, in relation to figure-ground organization, visual interpolation, and grouping.
Zygmunt Pizlo, Yunfeng Li, Tadamasa Sawada, and Robert M. Steinman
- Published in print:
- 2014
- Published Online:
- August 2014
- ISBN:
- 9780199922543
- eISBN:
- 9780190228385
- Item type:
- book
- Publisher:
- Oxford University Press
- DOI:
- 10.1093/acprof:oso/9780199922543.001.0001
- Subject:
- Psychology, Cognitive Neuroscience, Vision
This book explains why and how our visual perceptions are veridical; how they can provide us with an accurate representation of the world “out there.” It explains how this computationally difficult ...
More
This book explains why and how our visual perceptions are veridical; how they can provide us with an accurate representation of the world “out there.” It explains how this computationally difficult problem was solved by describing how the authors built a machine (a computational model) that sees very much as we do. This has never been done before and nothing remotely like it is available anywhere else. Doing it required a “paradigm shift,” an entirely new way of thinking about visual perception, one that is quite unlike any that has been considered up to now. The book, despite its scientific sophistication, is accessible to a very wide audience because each issue covered in the text is discussed twice, once for the “intuitive” reader and once for the “technical” reader. No equations are included in this book, but technical readers can find them in the authors’ published papers. The book, which contains many helpful demos, tells the story of how the machine was developed and what drove the ideas needed to make it work. This makes it an interesting, even gripping, read. The machine, explained clearly in this book, could have enormous practical and scientific, as well as social/artistic consequences. This book combines a new computational theory of shape perception with an account of the history of the theory's discovery. It tells this story together with all relevant background information including criticisms of it and of opposing theories. This mixture is an unusual way to present a major scientific achievement, but it not only works, it also makes for an exciting read.Less
This book explains why and how our visual perceptions are veridical; how they can provide us with an accurate representation of the world “out there.” It explains how this computationally difficult problem was solved by describing how the authors built a machine (a computational model) that sees very much as we do. This has never been done before and nothing remotely like it is available anywhere else. Doing it required a “paradigm shift,” an entirely new way of thinking about visual perception, one that is quite unlike any that has been considered up to now. The book, despite its scientific sophistication, is accessible to a very wide audience because each issue covered in the text is discussed twice, once for the “intuitive” reader and once for the “technical” reader. No equations are included in this book, but technical readers can find them in the authors’ published papers. The book, which contains many helpful demos, tells the story of how the machine was developed and what drove the ideas needed to make it work. This makes it an interesting, even gripping, read. The machine, explained clearly in this book, could have enormous practical and scientific, as well as social/artistic consequences. This book combines a new computational theory of shape perception with an account of the history of the theory's discovery. It tells this story together with all relevant background information including criticisms of it and of opposing theories. This mixture is an unusual way to present a major scientific achievement, but it not only works, it also makes for an exciting read.
Zygmunt Pizlo, Yunfeng Li, Tadamasa Sawada, and Robert M. Steinman
- Published in print:
- 2014
- Published Online:
- August 2014
- ISBN:
- 9780199922543
- eISBN:
- 9780190228385
- Item type:
- chapter
- Publisher:
- Oxford University Press
- DOI:
- 10.1093/acprof:oso/9780199922543.003.0006
- Subject:
- Psychology, Cognitive Neuroscience, Vision
This chapter explains how the front-end of our seeing machine works. This is often described as solving the “Figure-Ground-Organization” problem (FGO). FGO had proven to be the most difficult problem ...
More
This chapter explains how the front-end of our seeing machine works. This is often described as solving the “Figure-Ground-Organization” problem (FGO). FGO had proven to be the most difficult problem we faced when we began to fashion our machine. This first step in visual perception requires finding: (i) the ground surface, (ii) the objects standing on this ground, (iii) how many objects there are, (iv) where they are, and (v) how large they are. All of these operations are part of FGO. This information is sufficient for planning and executing the majority of our visually-guided navigations. Up to this point in our work, our computational models had required human intervention at the front end. Specifically, the human had to do three things: (i) locate where the objects are in the 2D image, (ii) mark their contours in the 2D image, and (iii) indicate their relationship. In this chapter, we explain how we elaborated on our model so it could do all of these things on its own. Human intervention is no longer needed.Less
This chapter explains how the front-end of our seeing machine works. This is often described as solving the “Figure-Ground-Organization” problem (FGO). FGO had proven to be the most difficult problem we faced when we began to fashion our machine. This first step in visual perception requires finding: (i) the ground surface, (ii) the objects standing on this ground, (iii) how many objects there are, (iv) where they are, and (v) how large they are. All of these operations are part of FGO. This information is sufficient for planning and executing the majority of our visually-guided navigations. Up to this point in our work, our computational models had required human intervention at the front end. Specifically, the human had to do three things: (i) locate where the objects are in the 2D image, (ii) mark their contours in the 2D image, and (iii) indicate their relationship. In this chapter, we explain how we elaborated on our model so it could do all of these things on its own. Human intervention is no longer needed.
Zygmunt Pizlo, Yunfeng Li, Tadamasa Sawada, and Robert M. Steinman
- Published in print:
- 2014
- Published Online:
- August 2014
- ISBN:
- 9780199922543
- eISBN:
- 9780190228385
- Item type:
- chapter
- Publisher:
- Oxford University Press
- DOI:
- 10.1093/acprof:oso/9780199922543.003.0007
- Subject:
- Psychology, Cognitive Neuroscience, Vision
This chapter first summarizes the most significant contributions of our work and then describes the three tasks that remain to be done, the tasks we are working on now. The first is to elaborate our ...
More
This chapter first summarizes the most significant contributions of our work and then describes the three tasks that remain to be done, the tasks we are working on now. The first is to elaborate our model to make it able to deal with dynamical environments in real-time. Humans do this all the time. They can also establish FGO from a single 2D image. Our machine, as built now, needs two. Adding this ability will make it able to process visual information faster and more reliably than it can now. We also are looking for a way to use symmetry to explain the human’s ability to recognize objects very quickly. This has never been tried before and we are optimistic about its potential for doing this. With these additions to our machine, if it works as well as our visual system, we think that it will finally be time to claim that our understanding of vision “must be nearly perfect” (Miller et al., 1960).Less
This chapter first summarizes the most significant contributions of our work and then describes the three tasks that remain to be done, the tasks we are working on now. The first is to elaborate our model to make it able to deal with dynamical environments in real-time. Humans do this all the time. They can also establish FGO from a single 2D image. Our machine, as built now, needs two. Adding this ability will make it able to process visual information faster and more reliably than it can now. We also are looking for a way to use symmetry to explain the human’s ability to recognize objects very quickly. This has never been tried before and we are optimistic about its potential for doing this. With these additions to our machine, if it works as well as our visual system, we think that it will finally be time to claim that our understanding of vision “must be nearly perfect” (Miller et al., 1960).
