Customer Reviews

Vision Science: Photons to Phenomenology

5 star:		(4)
4 star:		(2)
3 star:		(0)
2 star:		(0)
1 star:		(0)

 

 

This is an excellent book! Steve Palmer is perhaps best known for his work on perceptual grouping and perceptual organization. With this text, however, Palmer proves that his interests extend beyond these boundaries into the domain of perception as a whole. Palmer also demonstrates that his interests are not confined by one methodology either. Presenting findings from a number of perspectives is one of the things that excites Palmer the most, and he does it like no other. By making use of relevant research in neuroscience, psychology, computer vision and linguistics, to name a few, Palmer develops a unified text for the emerging domain of "Vision Science", a subfield of the larger interdisciplinary enterprise of Cognitive Science.

Palmer's book differs from other books on visual perception in three major ways. First, Palmer introduces the major theoretical perspectives to visual perception--inferential, ecological and computational-- early in the text and then places empirical findings throughout the text in the context of these perspectives. Second, Palmer presents findings from a number of disciplines in an integrated fashion. As opposed to having separate sections for neuroscience, computer vision and perceptual development, for example, Palmer presents research from multiple disciplines as it relates to relevant areas of visual perception, such as perceptual organization or object recognition. Third, and perhaps most importantly, Palmer resists the temptation to dichotomize. The discussions of the literature are sophisticated, presenting both the pros and the cons of different approaches to phenomena in perception, even venturing to propose novel theoretical syntheses at various points in the book.

For anyone who is interested in visual perception, neuroscience, computer vision, or just Cognitive Science in general, this is a book that you must have on your book shelf.

As an interested academic in a completely unrelated field (chemistry) Palmer's book was only the second cognitive psychology text that I had ever read. It was an "eye opener." There should be awards given to authors who commit themselves and succeed at the task of what Palmer has done here. The book was comprehensive and didn't pull any punches, but was still very readable. The quality of the writing and organization leads me to assume that the man is a gifted teacher as well. The layout, glossary, index, and organization of the text were clearly constructed with the reader in mind. Five star reviews at Amazon.com should be reserved for books of this quality.

Stephen Palmer has written a marvelous book. Its well organized and written. It is suitable as a reference & text for those beginning and advancing to higher levels in Vision Science. The pictures, diagrams, graphs, charts, photos, and outlines are well placed and explained in the body of the text. Because Dr. Palmer is first a Neuropsychologist, the approach is more geared toward psychological mechanisms & psychophysics. The emphasis is less on Biological Approaches to Visual Function. This & Chalupa's 2 Volume Set should get novice to intermediate scientists going further in Vision Science!

Book Review
"Vision Science: Photons to Phenomenology" by Stephen E Palmer, MIT Press, Cambridge, Massachusetts, 1999.

At 810 pages, this is a phenomenally encyclopaedic book. However, due justice to the grand title - "Photons to Phenomenology" - would by necessity take several volumes this size. I imagine the reader who scans the library shelf, licking their lips for a discourse about photons would be rather disappointed in this tome. On the other hand, the reader seeking a coherent account of current ideas concerning the phenomenon of visual perception will undoubtedly be delighted. So, despite being initially turned off by the now-hackneyed "X to Y"-style title (we have recently had "Pigments to Perception" and "Genes to Consciousness"... what next, "Molecules to Meaning" or perhaps "Big-bang to Business-bust"?), let me say why I became so fond of this book.

The stated aim is to present as full a view of Vision Science as possible, with offerings coming from the many diverse disciplines which claim an interest in the field: physiology, psychology, neuro-anatomy, computational neuroscience and psychophysics, to name a few. A tempting way to cover such a broad scientific landscape would be to maintain this convenient parcellation, and present chapters describing separately the findings of each approach. However the author avoids this temptation and instead undertakes the difficult task of integrating, as much as possible, pertinent information arising from different disciplines. Here Palmer has done more than a good job - he has managed to discover novel links between views offered by different disciplines and emerges with his own solid and internally consistent vision of Vision Science, which he then espouses deftly. Because of Palmer's background as a psychologist in the pursuit of understanding vision, this view is unmistakably biased towards the psychological, but it remains truly erudite and certainly far-reaching. The internal consistency of the resulting view is the great achievement of the book; in no place are the links between approaches overstated, but when available they are always intelligently described. Such a result would have been impossible to accomplish in any edited collection of chapters by individual experts in each separate 'branch' of Vision Science.

To facilitate this synthesis of Vision Science as a complete entity deriving from conjoined multidisciplinary roots, the book is presented in three broad parts. The first part, Foundations, includes an introduction chapter that defines visual perception and outlines the optical information that must be captured by the visual system in order to allow us to see. It is here that a brief overview of basic optics and the physiology of the eye and retina are covered. The second chapter discusses in broad-brush strokes the theories that have been proposed over recent history to account for how we see. Even though the goal is to achieve a synthesis of the many different ways of thinking about the problem of vision, Palmer's account of phenomena throughout the book is heavily referenced to particular theoretical underpinnings - the author takes pains to remind the reader of where different ideas come from. Structularlism, Gestaltism, Ecological Optics and Constructivism are introduced, followed by a treatment of where the "new" approach to vision, that of treating vision as essentially a computational problem, has its origins with a brief history of Information Processing Theory and its implications. Chapter three presents the case of colour and lightness perception as a microcosm of the larger problem of vision science; one which has benefited much from a multidisciplinary approach.

The second part of the book is called Spatial Vision, and includes six chapters describing different sequential elements of the process in which visual information is first extracted from the retinal image, and then used to inform the viewer about various aspects of the observed scene. Chapter 4 deals with what types of visual information are available in the retinal image, and the sorts of processing that must occur in the early parts of the visual system so that this information may be efficiently delivered to 'higher' areas - presumably other bits of the brain. This includes a description of the statistical nature of real-world images and how such knowledge may lead to 'reverse-engineering' the strategies employed by the visual system to cope with the mass of visual information continually raining down onto the retinal surface. Chapter five discusses the problem of how depth information in particular is extracted from the retinal images to produce solid perception of a 3D world. We are told how depth information may arise from various information sources: from ocular muscles, from retinal disparity (stereopsis), from motion, from previous knowledge of pictorial relationships and from shading cues. Chapter six informs about the problem of perceptual organisation of the visual world. Older and newer concepts of grouping principles are discussed, followed by newer theories of region analysis and visual interpolation. The phenomenon of multi-stability is offered as a tool for investigating how it is that the visual world comes to be perceptually organised, and connectionist network modelling is described as a way of investigating possible neural arrangements to account for the phenomenon. In chapter seven, the constancy in which objects and their parts are perceived is discussed. This includes size, shape, orientation and position constancy, and also what happens when these go awry - the basis for many well-known visual illusions. Chapters eight and nine cover increasingly more esoteric aspects of visual perception, describing the strengths and weaknesses of current and older theories of shape representation, and the perception of function and category respectively.

The third part of the book is called Visual Dynamics, and covers those aspects of visual perception that change over time. The best example of this is of course motion perception, which is the subject of Chapter 10. Motion perception arising from retinal image motion is covered from a computational point of view, quickly followed by discussion of the physiological mechanisms that are thought to underlie our ability to make sense of the shifting retinal image. Object motion, self-motion and optic flow are sub-topics further considered in follow-up sections. Chapter 11 attends to the problem of visual attention, and in particular how objects are selected for further consideration by the visual system and how eye movements are generated and made to enable such scrutiny. Theories of spatial attention including the "spotlight" and "zoom-lens" metaphors make thought-provoking reading. Chapter 12 covers visual memory and the ability of observers to recall the past by way of imagery. Both short-term and long-term memory theories and experiments are discussed. The final chapter attempts to deal with philosophically challenging concepts of visual awareness from a neuro-psychological viewpoint. It is wonderful to learn that advances are actually being made in this area, and that theories of consciousness are now being seriously explored by scientists and philosophers united.

This description of a tome is itself becoming somewhat heavy. Lest it remain an epitome, I had better wrap up. In summary I hope I have conveyed my admiration for this book. While I would advocate serious students of vision science purchase their own copy, I would not do so for all practitioners of clinical science unless they had a particular penchant for good writing (and lots of it!) in the genre of philosophical and theoretical science. Nevertheless, do borrow it!

- Andrew Metha
27 Aug 2002.

Palmer has written an outstanding book for the technically trained reader, even if that reader has a casual interest in understanding the emergence of of human cognition from the physics of vision, to the function of sight.
Absent from the book is a cogent presentation of aesthetics of seeing. For that the interested reader should see John Gage's Color and Meaning.

This book covers neuro, behavioral, computer science, almost everything about vision science, and very organized. at the bottom line, this book can be a good reference for vision science.