Hands-On Generative Adversarial Networks with Keras

GenerativeAdversarialNetworks(GANs)haverevolutionizedthefieldsofmachinelearninganddeeplearning.ThisbookwillbeyourfirststeptowardsunderstandingGANarchitecturesandtacklingthechallengesinvolvedintrainingthem.Thisbookopenswithanintroductiontodeeplearningandgenerativemodels,andtheirapplicationsinartificialintelligence(AI).Youwillthenlearnhowtobuild,evaluate,andimproveyourfirstGANwiththehelpofeasy-to-followexamples.ThenextfewchapterswillguideyouthroughtrainingaGANmodeltoproduceandimprovehigh-resolutionimages.YouwillalsolearnhowtoimplementconditionalGANsthatgiveyoutheabilitytocontrolcharacteristicsofGANoutputs.YouwillbuildonyourknowledgefurtherbyexploringanewtrainingmethodologyforprogressivegrowingofGANs.Movingon,you'llgaininsightsintostate-of-the-artmodelsinimagesynthesis,speechenhancement,andnaturallanguagegenerationusingGANs.Inadditiontothis,you'llbeabletoidentifyGANsampleswithTequilaGAN.Bytheendofthisbook,youwillbewell-versedwiththelatestadvancementsintheGANframeworkusingvariousexamplesanddatasets,andyouwillhavetheskillsyouneedtoimplementGANarchitecturesforseveraltasksanddomains,includingcomputervision,naturallanguageprocessing(NLP),andaudioprocessing.ForewordbyTing-ChunWang,SeniorResearchScientist,NVIDIA