Unveiling Functions of the Visual Cortex Using Task-specific Deep Neural Networks PDF Download

Author: Kshitij Dwivedi
Publisher:
ISBN:
Category :
Languages : en
Pages : 0

View

Book Description

Author: Anh M. Nguyen
Publisher:
ISBN: 9780355324389
Category : Artificial intelligence
Languages : en
Pages : 226

View

Book Description
Deep Learning, a type of Artificial Intelligence, is transforming many industries including transportation, health care and mobile computing. The main actors behind deep learning are deep neural networks (DNNs). These artificial brains have demonstrated impressive performance on many challenging tasks such as synthesizing and recognizing speech, driving cars, and even detecting cancer from medical scans. Given their excellent performance and widespread applications in everyday life, it is important to understand: (1) how DNNs function internally; (2) why they perform so well; and (3) when they fail. Answering these questions would allow end-users (e.g. medical doctors harnessing deep learning to assist them in diagnosis) to gain deeper insights into how these models behave, and therefore more confidence in utilizing the technology in important real-world applications. Artificial neural networks traditionally had been treated as black boxes—little was known about how they arrive at a decision when an input is present. Similarly, in neuroscience, understanding how biological brains work has also been a long-standing quest. Neuroscientists have discovered neurons in human brains that selectively fire in response to specific, abstract concepts such as Halle Berry or Bill Clinton, informing the discussion of whether learned neural codes are local or distributed. These neurons were identified by finding the preferred stimuli (here, images) that highly excite a specific neuron, which was accomplished by showing subjects many different images while recording a target neuron's activation. Inspired by such neuroscience techniques, my Ph.D. study produced a series of visualization methods that synthesize the preferred stimuli for each neuron in DNNs to shed more light into (1) the weaknesses of DNNs, which raise serious concerns about their widespread deployment in critical sectors of our economy and society; and (2) how DNNs function internally. Some of the notable findings are summarized as follows. First, DNNs are easily fooled in that it is possible to produce images that are visually unrecognizable to humans, but that state-of-the-art DNNs classify as familiar objects with near certainty confidence (i.e. labeling white-noise images as “school bus”). These images can be optimized to fool the DNN regardless of whether we treat the network as a white- or black-box (i.e. we have access to the network parameters or not). These results shed more light into the inner workings of DNNs and also question the security and reliability of deep learning applications. Second, our visualization methods reveal that DNNs can automatically learn a hierarchy of increasingly abstract features from the input space that are useful to solve a given task. In addition, we also found that neurons in DNNs are often multifaceted in that a single neuron fires for a variety of different input patterns (i.e. it is invariant to changes in the input). These observations align with the common wisdom previously established for both human visual cortex and DNNs. Lastly, many machine learning hobbyists and scientists have successfully applied our methods to visualize their own DNNs or even generate high-quality art images. We also turn the visualization frameworks into (1) an art generator algorithm, and (2) a state-of-the-art image generative model, making contributions to the fields of evolutionary computation and generative modeling, respectively.

Author: Stephane Molotchnikoff
Publisher: BoD – Books on Demand
ISBN: 9535107607
Category : Medical
Languages : en
Pages : 427

View

Book Description
The neurosciences have experienced tremendous and wonderful progress in many areas, and the spectrum encompassing the neurosciences is expansive. Suffice it to mention a few classical fields: electrophysiology, genetics, physics, computer sciences, and more recently, social and marketing neurosciences. Of course, this large growth resulted in the production of many books. Perhaps the visual system and the visual cortex were in the vanguard because most animals do not produce their own light and offer thus the invaluable advantage of allowing investigators to conduct experiments in full control of the stimulus. In addition, the fascinating evolution of scientific techniques, the immense productivity of recent research, and the ensuing literature make it virtually impossible to publish in a single volume all worthwhile work accomplished throughout the scientific world. The days when a single individual, as Diderot, could undertake the production of an encyclopedia are gone forever. Indeed most approaches to studying the nervous system are valid and neuroscientists produce an almost astronomical number of interesting data accompanied by extremely worthy hypotheses which in turn generate new ventures in search of brain functions. Yet, it is fully justified to make an encore and to publish a book dedicated to visual cortex and beyond. Many reasons validate a book assembling chapters written by active researchers. Each has the opportunity to bind together data and explore original ideas whose fate will not fall into the hands of uncompromising reviewers of traditional journals. This book focuses on the cerebral cortex with a large emphasis on vision. Yet it offers the reader diverse approaches employed to investigate the brain, for instance, computer simulation, cellular responses, or rivalry between various targets and goal directed actions. This volume thus covers a large spectrum of research even though it is impossible to include all topics in the extremely diverse field of neurosciences.

Author: David Ingle
Publisher: MIT Press (MA)
ISBN:
Category : Psychology
Languages : en
Pages : 870

View

Book Description
"Analysis of Visual Behavior" encompasses both theoretical and experimental research. It deals with the visual mechanisms of diverse vertebrate species from salamanders and toads to primates and humans and presents a stimulating interaction of the disciplines of anatomy, physiology, and behavioral science. Throughout, visual mechanisms are investigated from the point of view of the brain functioning at the organismic level, as opposed to the now more prevalent focus on the molecular and cellular levels. This approach allows researchers to deal with the patterns of visually guided behavior of animals in real-life situations.The twenty-six contributions in the book are divided among three sections: "Indentification and Localization Processes in Nonmammalian Vertebrates," introduced by David J. Ingle; "Visual Guidance of Motor Patterns: The Role of Visual Cortex and the Superior Colliculus," introduced by Melvyn A. Goodale; and "Recognition and Transfer Processes," introduced by Richard J. W. Mansfield.The editors are all university researchers in psychology: David J. Ingle at Brandeis, Melvyn A. Goodale at the University of Western Ontario, and Richard J. W. Mansfield at Harvard.

Author: Charles Legéndy
Publisher: Springer
ISBN: 9780387888484
Category : Medical
Languages : en
Pages : 226

View

Book Description
Dr. Charles Legéndy’s Circuits in the Brain: A Model of Shape Processing in the Primary Visual Cortex is published at a time marked by unprecedented advances in experimental brain research which are, however, not matched by similar advances in theoretical insight. For this reason, the timing is ideal for the appearance of Dr. Legéndy’s book, which undertakes to derive certain global features of the brain directly from the neurons. Circuits in the Brain, with its “relational firing” model of shape processing, includes a step-by-step development of a set of multi-neuronal networks for transmitting visual relations, using a strategy believed to be equally applicable to many aspects of brain function other than vision. The book contains a number of testable predictions at the neuronal level, some believed to be accessible to the techniques which have recently become available. With its novel approach and concrete references to anatomy and physiology, the monograph promises to open up entirely new avenues of brain research, and will be particularly useful to graduate students, academics, and researchers studying neuroscience and neurobiology. In addition, since Dr. Legéndy’s book succeeds in achieving a clean logical presentation without mathematics, and uses a bare minimum of technical terminology, it may also be enjoyed by non-scientists intrigued by the intellectual challenge of the elegant devices applied inside our brain. The book is uniquely self-contained; with more than 120 annotated illustrations it goes into full detail in describing all functional and theoretical concepts on which it builds.

Author: Jamell A. Dozier
Publisher:
ISBN:
Category :
Languages : en
Pages : 70

View

Book Description
Visual cognition has long been the subject of curiosity within the realm of deep learning. While much research has gone into the development of neural network models that can at times outperform humans, the underlying principles behind truly understanding visual concepts remain elusive. Utilizing a multitask learning paradigm, we first explore the capacity for networks to generalize to understand visual reasoning concepts. We introduce a simplified visual reasoning dataset to train several network architectures, including a recently proposed model built specifically for relational reasoning. We collect the best performing networks and view their behavior on a neuronal level: visualizing task selectivity through patterns of activations from each network layer. Finally, we adjust our focus to a simpler form of visual reasoning involving the extraction of single attributes from attribute compositions. Here, we are able to both visualize and quantify the neuron task selectivity that leads to generalization.

Author: Eshed Margalit
Publisher:
ISBN:
Category :
Languages : en
Pages : 0

View

Book Description
Humans have the remarkable capacity to recognize visual objects despite challenging variations in their pose, illumination, and context. This ability depends on the ventral visual stream, a series of cortical areas that progressively transforms the signal from the retina into representations of object category, location, color, texture, and size. Our understanding of the function and development of the ventral visual stream is anchored in the tight coupling between structure and function in the constituent cortical areas: in each area, neurons are arranged in the cortical sheet according to the visual features they respond most strongly to. In the earliest stage of the ventral visual stream neighboring neurons preferentially respond to edges of similar orientations and colors, whereas neurons toward the end of the ventral stream cluster together according to their preferred object category, e.g., faces, limbs, and places. Understanding the development and purpose of this functional organization requires the construction of detailed models whose predictions can be evaluated against neural measurements. In this dissertation, I present topographic deep convolutional neural networks (topographic DCNNs) as unifying models of neural structure and function throughout the ventral visual stream. Topographic DCNNs implement the simple hypothesis that functional organization in the visual cortex can be reproduced by optimizing the parameters of a neural network to perform a challenging visual task while keeping local populations of neurons correlated with one another. I find that topographic DCNNs are able to reproduce functional organization in both early and later stages of the ventral visual stream, that this brain-model correspondence is strongest for more biologically-plausible learning algorithms, and that topographic DCNNs can be used to predict how changes to visual inputs during development will affect cortical map formation. The success of topographic DCNNs in the prediction of the functional organization of the primate ventral visual stream implies the existence of simple unifying principles for the development of those regions, and serves as a foundation from which increasingly accurate models of visual processing can be constructed.

Author: Judith C. Peters
Publisher: Frontiers Media SA
ISBN: 2889198731
Category : Neurosciences. Biological psychiatry. Neuropsychiatry
Languages : en
Pages : 139

View

Book Description
The articles in this Research Topic provide a state-of-the-art overview of the current progress in integrating computational and empirical research on visual object recognition. Developments in this exciting multidisciplinary field have recently gained momentum: High performance computing enabled breakthroughs in computer vision and computational neuroscience. In parallel, innovative machine learning applications have recently become available for datamining the large-scale, high resolution brain data acquired with (ultra-high field) fMRI and dense multi-unit recordings. Finally, new techniques to integrate such rich simulated and empirical datasets for direct model testing could aid the development of a comprehensive brain model. We hope that this Research Topic contributes to these encouraging advances and inspires future research avenues in computational and empirical neuroscience.

Author: David Rose
Publisher: John Wiley & Sons
ISBN:
Category : Medical
Languages : en
Pages : 616

View

Book Description
A comprehensive and stimulating study which presents the views of 71 leading theorists on the underlying mechanisms and functions of the primary visual cortex.

Author: William D. Penny
Publisher: Elsevier
ISBN: 0080466508
Category : Psychology
Languages : en
Pages : 689

View

Book Description
In an age where the amount of data collected from brain imaging is increasing constantly, it is of critical importance to analyse those data within an accepted framework to ensure proper integration and comparison of the information collected. This book describes the ideas and procedures that underlie the analysis of signals produced by the brain. The aim is to understand how the brain works, in terms of its functional architecture and dynamics. This book provides the background and methodology for the analysis of all types of brain imaging data, from functional magnetic resonance imaging to magnetoencephalography. Critically, Statistical Parametric Mapping provides a widely accepted conceptual framework which allows treatment of all these different modalities. This rests on an understanding of the brain's functional anatomy and the way that measured signals are caused experimentally. The book takes the reader from the basic concepts underlying the analysis of neuroimaging data to cutting edge approaches that would be difficult to find in any other source. Critically, the material is presented in an incremental way so that the reader can understand the precedents for each new development. This book will be particularly useful to neuroscientists engaged in any form of brain mapping; who have to contend with the real-world problems of data analysis and understanding the techniques they are using. It is primarily a scientific treatment and a didactic introduction to the analysis of brain imaging data. It can be used as both a textbook for students and scientists starting to use the techniques, as well as a reference for practicing neuroscientists. The book also serves as a companion to the software packages that have been developed for brain imaging data analysis. - An essential reference and companion for users of the SPM software - Provides a complete description of the concepts and procedures entailed by the analysis of brain images - Offers full didactic treatment of the basic mathematics behind the analysis of brain imaging data - Stands as a compendium of all the advances in neuroimaging data analysis over the past decade - Adopts an easy to understand and incremental approach that takes the reader from basic statistics to state of the art approaches such as Variational Bayes - Structured treatment of data analysis issues that links different modalities and models - Includes a series of appendices and tutorial-style chapters that makes even the most sophisticated approaches accessible