Generative Adversarial Networks Cookbook
DevelopingGenerativeAdversarialNetworks(GANs)isacomplextask,anditisoftenhardtofindcodethatiseasytounderstand.ThisbookleadsyouthrougheightdifferentexamplesofmodernGANimplementations,includingCycleGAN,simGAN,DCGAN,and2Dimageto3Dmodelgeneration.EachchaptercontainsusefulrecipestobuildonacommonarchitectureinPython,TensorFlowandKerastoexploreincreasinglydifficultGANarchitecturesinaneasy-to-readformat.ThebookstartsbycoveringthedifferenttypesofGANarchitecturetohelpyouunderstandhowthemodelworks.ThisbookalsocontainsintuitiverecipestohelpyouworkwithusecasesinvolvingDCGAN,Pix2Pix,andsoon.Tounderstandthesecomplexapplications,youwilltakedifferentreal-worlddatasetsandputthemtouse.Bytheendofthisbook,youwillbeequippedtodealwiththechallengesandissuesthatyoumayfacewhileworkingwithGANmodels,thankstoeasy-to-followcodesolutionsthatyoucanimplementrightaway.
·4.7萬字
