Van Hemmen J.L., Cowan J.D., Domany E. (eds.) Models of Neural Networks IV. Early Vision and Attention

Springer, 2002. — 424 p.The fourth volume of the Physics of Neural Networks series.Models of Neural Networks I
Models of Neural Networks II. Temporal Aspects of Coding and Information Processing in Biological Systems
Models of Neural Networks III. Association, Generalization, and Representation
Models of Neural Networks IV. Early Vision and AttentionClose this book for a moment and look around you. You scan the scene by directing your attention, and gaze, at certain specific objects. Despite the background, you discern them. The process is partially intentional and partially preattentive. How all this can be done is described in the fourth volume of Models of Neural Networks devoted to Early Vision and Attention that you are holding in your hands. Early vision comprises the first stages of visual information processing. It is as such a scientific challenge whose clarification calls for a penetrating review. Here you see the result. The Heraeus Foundation (Hanau) is to be thanked for its support during the initial phase of this project.
John Hertz, who has extensive experience in both computational and experimental neuroscience, provides in "Neurons, Networks, and Cognition" a theoretical introduction to neural modeling. John Van Opstal explains in "The Gaze Control System" how the eye's gaze control is performed and presents a novel theoretical description incorporating recent experimental results. We then turn to the relay stations thereafter, the lateral geniculate nucleus (LGN) and the primary visual cortex. Their anatomy, physiology, functional relations, and ensuing response properties are carefully analyzed by Klaus Funke et al. in "Integrating Anatomy and Physiology of the Primary Visual Pathway: From LGN to Cortex", one of the most comprehensive reviews that is available at the moment.
How do we discern patterns? That is to say, how do we perform scene segmentation? It has been shown that this process is partially preattentive and, so to speak, done on the spot in the primary visual cortex. Reinhard Eckhorn explains the underlying "Neural Principles of Preattentive Scene Segmentation" while Esther Peterhans et al. analyze a neuronal model of "Figure-Ground Segregation and Brightness Perception at Illusory Contours".
Scene segmentation can also be performed by a feedback process that is called 'attention'. A glance suffices to convince every beholder that the eye catches megabytes of data. Through attention we reduce this data flood by singling out specific objects. Ernst Niebur et al. indicate how this can be done by "Controlling the Focus of Visual Selective Attention" while Julian Eggert and Leo van Hemmen elucidate the feedback mechanism proper in "Activity-Gating Attentional Networks".
Ever tried to smash a busy buzzing fly against the wall? Then you know how good it is in avoiding you. That is to say, you realize that also insects such as flies may perform highly efficient visual-information processing. In their essay "Timing and Counting Precision in the Blowfly Visual System" Rob de Ruyter van Steveninck and Bill Bialek explain how this is done in early vision and show what key role is played by spikes. Finally, Wolfgang Maass approaches "Paradigms for Computing with Spiking Neurons" from the point of view of a computer scientist who is concerned with biological information processing. Enjoy!Neurons, Networks, and Cognition: An Introduction to Neural Modeling
The Gaze Control System
Integrating Anatomy and Physiology of the Primary Visual Pathway: From LGN to Cortex
Neural Principles of Preattentive Scene Segmentation: Hints from Cortical Recordings, Related Models, and Perception
Figure-Ground Segregation and Brightness Perception at Illusory Contours: A Neuronal Model
Controlling the Focus of Visual Selective Attention
Activity-Gating Attentional Networks
Timing and Counting Precision in the Blowfly Visual System
Paradigms for Computing with Spiking Neurons