Generative Adversarial Networks Projects
GenerativeAdversarialNetworks(GANs)havethepotentialtobuildnext-generationmodels,astheycanmimicanydistributionofdata.Majorresearchanddevelopmentworkisbeingundertakeninthisfieldsinceitisoneoftherapidlygrowingareasofmachinelearning.Thisbookwilltestunsupervisedtechniquesfortrainingneuralnetworksasyoubuildsevenend-to-endprojectsintheGANdomain.GenerativeAdversarialNetworkProjectsbeginsbycoveringtheconcepts,tools,andlibrariesthatyouwillusetobuildefficientprojects.Youwillalsouseavarietyofdatasetsforthedifferentprojectscoveredinthebook.Thelevelofcomplexityoftheoperationsrequiredincreaseswitheverychapter,helpingyougettogripswithusingGANs.Youwillcoverpopularapproachessuchas3D-GAN,DCGAN,StackGAN,andCycleGAN,andyou’llgainanunderstandingofthearchitectureandfunctioningofgenerativemodelsthroughtheirpracticalimplementation.Bytheendofthisbook,youwillbereadytobuild,train,andoptimizeyourownend-to-endGANmodelsatworkorinyourownprojects.
·5.2萬字