Deep Learning with Microsoft Cognitive Toolkit
Published on 28. February 2018
Book
Paperback/Softback
336 pages
978-1-78712-832-3 (ISBN)
Description
Leverage the power of Deep learning with Microsoft's very own open source framework
About This Book
Build deployable solutions to tackle common deep learning problems.
Build high speed and efficient deep learning models using Microsoft Cognitive Toolkit.
Explore the various neural networks with the help of this comprehensive guide
Who This Book Is ForThis book is intended for data science professionals interested in deep learning and who would like to explore the new features of Microsoft CNTK. Basic Machine Learning and programming knowledge is assumed. Anyone looking for fetching deeper insights into their data with fast, open source tools will find this book quite helpful.
What You Will Learn
Learn basic concepts in deep learning
Know how to prepare data that can be consumed by CNTK for training
Discover how to train simple deep learning models with the Python API of CNTK
Evaluate trained CNTK models on new test data sets
Scale the execution of model training across multiple GPUs and multiple machines
Explore the common deep learning models for speech recognition, language understanding, image classification, etc.
Understand how to use CNTK to solve real-word problems such as face and emotion recognition, image segmentation, neural artistic style, image captioning, gaming, etc.
In DetailRight from setting up your neural network, this book will guide you to achive incredible computation speed with Microsoft Cognitive Tooklit. We will delve into machine learning aspects like speech comprehension, text to speech conversion, voice recognition, object recognition and many more. You will learn about the Network Description Language (NDL) and setting up the neural network. Further you will then explore the various deep learning architectures (CNNs, DNNs, RNNs etc) and how to use them to accelerate your computations. With practical examples, this book will teach you to work with time-series data and to create, train, and validate test sets. Later, you will understand how to work with larger datasets and dealing with heterogenous input data. You will also learn about the latest advances such as bindings with Python and C++.
By the end of this book, you will be able to migrate from other toolkits.
About This Book
Build deployable solutions to tackle common deep learning problems.
Build high speed and efficient deep learning models using Microsoft Cognitive Toolkit.
Explore the various neural networks with the help of this comprehensive guide
Who This Book Is ForThis book is intended for data science professionals interested in deep learning and who would like to explore the new features of Microsoft CNTK. Basic Machine Learning and programming knowledge is assumed. Anyone looking for fetching deeper insights into their data with fast, open source tools will find this book quite helpful.
What You Will Learn
Learn basic concepts in deep learning
Know how to prepare data that can be consumed by CNTK for training
Discover how to train simple deep learning models with the Python API of CNTK
Evaluate trained CNTK models on new test data sets
Scale the execution of model training across multiple GPUs and multiple machines
Explore the common deep learning models for speech recognition, language understanding, image classification, etc.
Understand how to use CNTK to solve real-word problems such as face and emotion recognition, image segmentation, neural artistic style, image captioning, gaming, etc.
In DetailRight from setting up your neural network, this book will guide you to achive incredible computation speed with Microsoft Cognitive Tooklit. We will delve into machine learning aspects like speech comprehension, text to speech conversion, voice recognition, object recognition and many more. You will learn about the Network Description Language (NDL) and setting up the neural network. Further you will then explore the various deep learning architectures (CNNs, DNNs, RNNs etc) and how to use them to accelerate your computations. With practical examples, this book will teach you to work with time-series data and to create, train, and validate test sets. Later, you will understand how to work with larger datasets and dealing with heterogenous input data. You will also learn about the latest advances such as bindings with Python and C++.
By the end of this book, you will be able to migrate from other toolkits.
More details
Language
English
Place of publication
Birmingham
United Kingdom
Dimensions
Height: 235 mm
Width: 191 mm
ISBN-13
978-1-78712-832-3 (9781787128323)
Copyright in bibliographic data and cover images is held by Nielsen Book Services Limited or by the publishers or by their respective licensors: all rights reserved.
Schweitzer Classification
Persons
Emad Barsoum is a Principal SDE and Applied Researcher in the Advance Technology Group at Microsoft Research. His current research focus is on computer vision and deep learning algorithms, especially in the areas of emotion recognition, activity detection/recognition and caption generation from video and image signals.
Emad was one of the core developer and researcher behind the Emotion Recognition Algorithm used in MS Cognitive Service for both still image and video. Before that, he was one of the main Architects for NUI API in Xbox One, and developed the depth reconstruction algorithm for Kinect 2 in collaboration with Microsoft Research. Furthermore, he helped developing and shipping fitness algorithm from skeleton and real-time image segmentation algorithm based on geodesic transform for Kinect 2.
Emad received his M.S. degree in EECS from UC. Irvine 2004, focused on Computer Vision and Virtual Reality, and B.S. degree in EECS from Ain Shams University 2000. He is currently pursuing a doctorate degree in the Computer Science department of Columbia University. William Darling is an Applied Scientist working in Microsoft's AI and Research team. He received an H.B.Sc. in Physics and Computer Science from Trent University in 2004, and LL.B. and B.C.L. degrees from McGill University in 2007. After being called to the bar of the Law Society of Upper Canada and working in intellectual property law in Toronto, he returned to academia and completed a Ph.D. focused on Machine Learning and Topic Modeling from the University of Guelph in 2012. He is currently working on the AI Platform Team for Bing with a focus on deep recurrent neural network models and the Microsoft Cognitive Toolkit. He has published and commercialized technology in the areas of NLP, machine learning, and information retrieval. Willi Richert is a Senior Applied Scientist on Microsoft's AI and Research team. He received a PhD in Machine Learning/Robotics in 2009 from Paderborn University, Germany, where he used reinforcement learning, hidden Markov models, and Bayesian networks to let heterogeneous robots learn by imitation. Currently, he is involved in improving the Microsoft Cognitive Toolkit and applying it to Bing workloads. He is the coauthor of "Building Machine Learning Systems with Python", which is now in its 2nd edition. Frank Seide, a native of Hamburg, Germany, is a Principal Researcher at Microsoft Research, and an architect of Microsoft's Cognitive Toolkit for deep learning. His current research focus is on deep neural networks for conversational speech recognition; together with co-author Dong Yu, he was first to show the effectiveness of deep neural networks for recognition of conversational speech. Throughout his career, he has been interested in and worked on a broad range of topics and components of automatic speech recognition, including spoken-dialogue systems, recognition of Mandarin Chinese, and, particularly, large-vocabulary recognition of conversational speech with application to audio indexing, transcription, and speech-to-speech translation.
In 1993, Frank received a Master degree in electrical engineering from the University of Technology of Hamburg-Harburg, Germany, and joined the speech research group of Philips Research in Aachen, Germany, to work on spoken-dialogue systems. He then transferred to Taiwan as one of the founding members of Philips Research East-Asia, Taipei, to lead a research project on Mandarin speech recognition. In June 2001, he joined the speech group at Microsoft Research Asia, Beijing, initially as a Researcher, since 2003 as Project Leader for offline speech applications, and since October 2006 as Research Manager. In 2014, Frank joined the Speech & Dialogue group at MSR Redmond as a Principal Researcher.
Cha Zhang is a Principal Researcher in the Advanced Technology Group at Microsoft Research. He received the B.S. and M.S. degrees from Tsinghua University, Beijing, China in 1998 and 2000, respectively, both in Electronic Engineering, and the Ph.D. degree in Electrical and Computer Engineering from Carnegie Mellon University, in 2004. His current research focuses on applying various audio/image/video processing and machine learning techniques to multimedia applications. Dr. Zhang has published more than 80 technical papers and holds 30+ U.S. patents. He won the best paper award at ICME 2007, the top 10% award at MMSP 2009, and the best student paper award at ICME 2010. He was the Program Co-Chair for VCIP 2012, and the General Co-Chair for ICME 2016. He serves as Area Chair for CVPR 2017. He currently is an Associate Editor for IEEE Trans. on Circuits and Systems for Video Technology, and IEEE Trans. on Multimedia.
Emad was one of the core developer and researcher behind the Emotion Recognition Algorithm used in MS Cognitive Service for both still image and video. Before that, he was one of the main Architects for NUI API in Xbox One, and developed the depth reconstruction algorithm for Kinect 2 in collaboration with Microsoft Research. Furthermore, he helped developing and shipping fitness algorithm from skeleton and real-time image segmentation algorithm based on geodesic transform for Kinect 2.
Emad received his M.S. degree in EECS from UC. Irvine 2004, focused on Computer Vision and Virtual Reality, and B.S. degree in EECS from Ain Shams University 2000. He is currently pursuing a doctorate degree in the Computer Science department of Columbia University. William Darling is an Applied Scientist working in Microsoft's AI and Research team. He received an H.B.Sc. in Physics and Computer Science from Trent University in 2004, and LL.B. and B.C.L. degrees from McGill University in 2007. After being called to the bar of the Law Society of Upper Canada and working in intellectual property law in Toronto, he returned to academia and completed a Ph.D. focused on Machine Learning and Topic Modeling from the University of Guelph in 2012. He is currently working on the AI Platform Team for Bing with a focus on deep recurrent neural network models and the Microsoft Cognitive Toolkit. He has published and commercialized technology in the areas of NLP, machine learning, and information retrieval. Willi Richert is a Senior Applied Scientist on Microsoft's AI and Research team. He received a PhD in Machine Learning/Robotics in 2009 from Paderborn University, Germany, where he used reinforcement learning, hidden Markov models, and Bayesian networks to let heterogeneous robots learn by imitation. Currently, he is involved in improving the Microsoft Cognitive Toolkit and applying it to Bing workloads. He is the coauthor of "Building Machine Learning Systems with Python", which is now in its 2nd edition. Frank Seide, a native of Hamburg, Germany, is a Principal Researcher at Microsoft Research, and an architect of Microsoft's Cognitive Toolkit for deep learning. His current research focus is on deep neural networks for conversational speech recognition; together with co-author Dong Yu, he was first to show the effectiveness of deep neural networks for recognition of conversational speech. Throughout his career, he has been interested in and worked on a broad range of topics and components of automatic speech recognition, including spoken-dialogue systems, recognition of Mandarin Chinese, and, particularly, large-vocabulary recognition of conversational speech with application to audio indexing, transcription, and speech-to-speech translation.
In 1993, Frank received a Master degree in electrical engineering from the University of Technology of Hamburg-Harburg, Germany, and joined the speech research group of Philips Research in Aachen, Germany, to work on spoken-dialogue systems. He then transferred to Taiwan as one of the founding members of Philips Research East-Asia, Taipei, to lead a research project on Mandarin speech recognition. In June 2001, he joined the speech group at Microsoft Research Asia, Beijing, initially as a Researcher, since 2003 as Project Leader for offline speech applications, and since October 2006 as Research Manager. In 2014, Frank joined the Speech & Dialogue group at MSR Redmond as a Principal Researcher.
Cha Zhang is a Principal Researcher in the Advanced Technology Group at Microsoft Research. He received the B.S. and M.S. degrees from Tsinghua University, Beijing, China in 1998 and 2000, respectively, both in Electronic Engineering, and the Ph.D. degree in Electrical and Computer Engineering from Carnegie Mellon University, in 2004. His current research focuses on applying various audio/image/video processing and machine learning techniques to multimedia applications. Dr. Zhang has published more than 80 technical papers and holds 30+ U.S. patents. He won the best paper award at ICME 2007, the top 10% award at MMSP 2009, and the best student paper award at ICME 2010. He was the Program Co-Chair for VCIP 2012, and the General Co-Chair for ICME 2016. He serves as Area Chair for CVPR 2017. He currently is an Associate Editor for IEEE Trans. on Circuits and Systems for Video Technology, and IEEE Trans. on Multimedia.