內容簡介
內容簡介 A gentle introduction to Generative Adversarial Networks, and a practical step-by-step tutorial on making your own with PyTorch. This beginner-friendly guide will give you hands-on experience: learning PyTorch basicsdeveloping your first PyTorch neural networkexploring neural network refinements to improve performanceintroduce CUDA GPU accelerationIt will introduce GANs, one of the most exciting areas of machine learning: introducing the concept step-by-step, in plain Englishcoding the simplest GAN to develop a good workflowgrowing our confidence with an MNIST GANprogressing to develop a GAN to generate full-colour human facesexperiencing how GANs fail, exploring remedies and improving GAN performance and stabilityBeyond the very basics, readers can explore more sophisticated GANs: convolutional GANs for generated higher quality imagesconditional GANs for generated images of a desired classThe appendices will be useful for students of machine learning as they explain themes often skipped over in many courses: calculating ideal loss values for balanced GANsprobability distributions and sampling them to create imagescarefully chosen examples illustrating how convolutions worka brief explanation of why gradient descent isn't suited to adversarial machine learningAll code is available publicly as open source on github.
