Dynamic Neural Field Theory for Motion Perception provides a new theoretical framework that permits a systematic analysis of the dynamic properties of motion perception. This framework uses dynamic neural fields as a key mathematical concept. The author demonstrates how neural fields can be applied for the analysis of perceptual phenomena and its underlying neural processes. Also, similar principles form a basis for the design of computer vision systems as well as the design of artificially behaving systems. The book discusses in detail the application of this theoretical approach to motion perception and will be of great interest to researchers in vision science, psychophysics, and biological visual systems.
Neural field theory has a long-standing tradition in the mathematical and computational neurosciences. Beginning almost 50 years ago with seminal work by Griffiths and culminating in the 1970ties with the models of Wilson and Cowan, Nunez and Amari, this important research area experienced a renaissance during the 1990ties by the groups of Ermentrout, Robinson, Bressloff, Wright and Haken. Since then, much progress has been made in both, the development of mathematical and numerical techniques and in physiological refinement und understanding. In contrast to large-scale neural network models described by huge connectivity matrices that are computationally expensive in numerical simulations, neural field models described by connectivity kernels allow for analytical treatment by means of methods from functional analysis. Thus, a number of rigorous results on the existence of bump and wave solutions or on inverse kernel construction problems are nowadays available. Moreover, neural fields provide an important interface for the coupling of neural activity to experimentally observable data, such as the electroencephalogram (EEG) or functional magnetic resonance imaging (fMRI). And finally, neural fields over rather abstract feature spaces, also called dynamic fields, found successful applications in the cognitive sciences and in robotics. Up to now, research results in neural field theory have been disseminated across a number of distinct journals from mathematics, computational neuroscience, biophysics, cognitive science and others. There is no comprehensive collection of results or reviews available yet. With our proposed book Neural Field Theory, we aim at filling this gap in the market. We received consent from some of the leading scientists in the field, who are willing to write contributions for the book, among them are two of the founding-fathers of neural field theory: Shun-ichi Amari and Jack Cowan.
This volume includes papers originally presented at the 11th annual Computational Neuroscience Meeting (CNS 02) held in July 2002 at the Congress Plaza Hotel & Convention Center in Chicago, Illinois, USA. The CNS meetings bring together computational neuroscientists representing many different fields and backgrounds as well as many different experimental preparations and theoretical approaches. The papers published here range from pure experimental neurobiology, to neuro-ethology, mathematics, physics, and engineering. In all cases the research described is focused on understanding how nervous systems compute. The actual subjects of the research include a highly diverse number of preparations, modeling approaches and analysis techniques. Accordingly, this volume reflects the breadth and depth of current research in computational neuroscience taking place throughout the world.
This book details a model of consciousness supported by scientific experimental data from the human brain. It presents how the Corollary Discharge of Attention Movement (CODAM) neural network model allows for a scientific understanding of consciousness as well as provides a solution to the Mind-Body problem. The book provides readers with a general approach to consciousness that is powerful enough to lead to the inner self and its ramifications for the vast range of human experiences. It also offers an approach to the evolution of human consciousness and features chapters on mental disease (especially schizophrenia) and on meditative states (including drug-induced states of mind). Solving the Mind-Body Problem bridges the gap that exists between philosophers of mind and the neuroscience community, allowing the enormous weight of theorizing on the nature of mind to be brought to earth and put under the probing gaze of the scientific facts of life and mind.
For more than 30 years, the visual cortex has been the source of new theories and ideas about how the brain processes information. The visual cortex is easily accessible through a variety of recording and imagining techniques and allows mapping of high level behavior relatively directly to neural mechanisms. Understanding the computations in the visual cortex is therefore an important step toward a general theory of computational brain theory.
The job of any sensory system is to create objects in the world out of the incoming proximal stimulus energy. The energy is neutral; it does not specify the objects itself. Thus, sensory systems must abstract the energy that does specify objects and differentiate it from the noise energy. The perceptual variables that specify objects for both listening and looking become those of contrast and correlated change across space and time, so that perceiving occurs at several spatial and temporal scales in parallel. Given that the perceptual goals and perceptual variables are equivalent, the rules of perceiving will be the same for all senses. The goal of this book is to describe these conceptual similarities and differences between hearing and seeing. Although it is mathematical and conceptually analytical, the book does not make explicit use of advanced mathematical concepts. Each chapter combines information on hearing and seeing, and gives a detailed treatment of a small number of topics. The first three chapters present introductory information, including properties of auditory and visual worlds, how receptive fields are organized to pick out those properties, and whether the receptive fields are optimized to pick up the structure of the sensory world. Each subsequent chapter considers one type of perceptual element: texture, motion, contrast and noise, color, timbre, and object segmentation. Each type of perceptual situation is described as a problem of discovering the correlated energy, and the research presented focuses on how humans manage to perceive given the complicated set of skills required. This book is intended for use in upper-division undergraduate courses in perception and sensation, cognitive psychology, and neuroscience. It will fill the slot between textbooks that cover perception and sensory physiology and neuroscience, and more advanced monographs that cover one sense or topic in detail.
This book is about the dynamics of neural systems and should be suitable for those with a background in mathematics, physics, or engineering who want to see how their knowledge and skill sets can be applied in a neurobiological context. No prior knowledge of neuroscience is assumed, nor is advanced understanding of all aspects of applied mathematics! Rather, models and methods are introduced in the context of a typical neural phenomenon and a narrative developed that will allow the reader to test their understanding by tackling a set of mathematical problems at the end of each chapter. The emphasis is on mathematical- as opposed to computational-neuroscience, though stresses calculation above theorem and proof. The book presents necessary mathematical material in a digestible and compact form when required for specific topics. The book has nine chapters, progressing from the cell to the tissue, and an extensive set of references. It includes Markov chain models for ions, differential equations for single neuron models, idealised phenomenological models, phase oscillator networks, spiking networks, and integro-differential equations for large scale brain activity, with delays and stochasticity thrown in for good measure. One common methodological element that arises throughout the book is the use of techniques from nonsmooth dynamical systems to form tractable models and make explicit progress in calculating solutions for rhythmic neural behaviour, synchrony, waves, patterns, and their stability. This book was written for those with an interest in applied mathematics seeking to expand their horizons to cover the dynamics of neural systems. It is suitable for a Masters level course or for postgraduate researchers starting in the field of mathematical neuroscience.
Human performance in visual perception by far exceeds the performance of contemporary computer vision systems. While humans are able to perceive their environment almost instantly and reliably under a wide range of conditions, computer vision systems work well only under controlled conditions in limited domains. This book sets out to reproduce the robustness and speed of human perception by proposing a hierarchical neural network architecture for iterative image interpretation. The proposed architecture can be trained using unsupervised and supervised learning techniques. Applications of the proposed architecture are illustrated using small networks. Furthermore, several larger networks were trained to perform various nontrivial computer vision tasks.
The book constitutes the proceedings of the 24th International Conference on Artificial Neural Networks, ICANN 2014, held in Hamburg, Germany, in September 2014. The 107 papers included in the proceedings were carefully reviewed and selected from 173 submissions. The focus of the papers is on following topics: recurrent networks; competitive learning and self-organisation; clustering and classification; trees and graphs; human-machine interaction; deep networks; theory; reinforcement learning and action; vision; supervised learning; dynamical models and time series; neuroscience; and applications.
