Ganguly Kuntal. Learning Generative Adversarial Networks

Packt Publishing, 2017. — 180 p. — ISBN: 1788396413.Build image generation and semi-supervised models using Generative Adversarial Networks.Generative models are gaining a lot of popularity among the data scientists, mainly because they facilitate the building of AI systems that consume raw data from a source and automatically builds an understanding of it. Unlike supervised learning methods, generative models do not require labeling of the data which makes it an interesting system to use. This book will help you to build and analyze the deep learning models and apply them to real-world problems. This book will help readers develop intelligent and creative application from a wide variety of datasets, mainly focusing on visuals or images.The book begins with the basics of generative models, as you get to know the theory behind Generative Adversarial Networks and its building blocks. This book will show you how you can overcome the problem of text to image synthesis with GANs, using libraries like Tensorflow, Keras and PyTorch. Transfering style from one domain to another becomes a headache when working with huge data sets. The author, using real-world examples, will show how you can overcome this. You will understand and train Generative Adversarial Networks and use them in a production environment and learn tips to use them effectively and accurately.What this book coversWhat you need for this book.Who this book is forConventions.
Reader feedback.
Customer support.
Introduction to Deep Learning
Evolution of deep learning.
Deconvolution or transpose convolution.Unsupervised Learning with GAN
Automating human tasks with deep neural networks.
Implementation of GAN.
Challenges of GAN models.
Improved training approaches and tips for GAN.Transfer Image Style Across Various Domains
Bridging the gap between supervised and unsupervised learning.
Introduction to Conditional GAN.
The training procedure of BEGAN.
Image to image style transfer with CycleGAN.Building Realistic Images from Your Text
Introduction to StackGAN.
Discovering cross-domain relationships with DiscoGAN.
Generating handbags from edges with PyTorch.
Gender transformation using PyTorch.
DiscoGAN versus CycleGAN.Using Various Generative Models to Generate Images
Introduction to Transfer Learning.
Large scale deep learning with Apache Spark.
Generating artistic hallucinated images using DeepDream.
Generating handwritten digits with VAE using TensorFlow.
A real world analogy of VAE.
A comparison of two generative models — GAN and VAE.Taking Machine Learning to Production
Building an image correction system using DCGAN.
Microservice architecture using containers.
Various approaches to deploying deep models.
Serving Keras-based deep models on Docker.
Deploying a deep model on the cloud with GKE.
Serverless image recognition with audio using AWS Lambda and Polly.