Kedis ku xidhan hagaha · Fudud Heerka

Kediska GAN-yada Shuruuda ah

Wuxuu hubiyaa sida agabka shuruuda ah u hago labada shabakadood ee GAN, sida takoorayaashu u qiimeeyaan lammaanaha is-bar-bar socda, iyo sida qaboojinta loo duro.

Wadooyinka haga ee la xidhiidhaGAN-yada shuruudaysan
Su'aal 1 ee 8

Waa maxay faraqa ugu weyn ee u dhexeeya GAN shuruudaysan iyo GAN caadiga ah?

Horudhac su'aal kasta8 su'aalo

Jawaabaha iyo sharraxaadaha waxay soo baxayaan markaad qaadanayso kediska.

  1. Waa maxay faraqa ugu weyn ee u dhexeeya GAN shuruudaysan iyo GAN caadiga ah?
    • Waxay quudisaa macluumaad dheeraad ah sida summada fasalka labada shabakadood si loo xakameeyo wax soo saarka
    • Waxay isticmaashaa laba matoor
    • Waxay meesha ka saartaa takoorka
    • Waxa kaliya oo ay ku shaqeysaa qoraalka, maaha sawirro
  2. CGAN lagu tabobaray nambaro calaamadeysan, maxaad sameyn kartaa oo uusan GAN caadi ah awoodin?
    • Orod iyada oo aan wax buuq ah gelin
    • Codso tiro gaar ah sida '7'
    • Gebi ahaanba ka gudub takoorka
    • Samee kaliya shanqadh madow iyo caddaan
  3. Maxay tahay in cGAN-ta takoorku hubiyo, ka baxsan in sawirku run u ekaado?
    • Qaabka faylka sawirka
    • xallinta ee megapixels
    • Haddii sawirku la mid yahay calaamadda la siiyay
    • Heerka waxbarasho ee tababarka
  4. Waa kuwee dariiqa caadiga ah ee lagu siiyo calaamada qaboojinta koronto-dhaliye?
    • Tirtiridda vectorka buuqa
    • Tababar aan lahayn wax calaamado ah
    • Isticmaalka dareeme jireed oo gooni ah
    • Isku-dubbaridka summada iyo gelinta dhawaaqa
  5. GAN-yada shardiga ah ayaa gundhig u ahaa awoodda dambe?
    • Qoraal-ilaa-sawir iyo jiil-sawir-sawir
    • CPUs aad u dheereya
    • Shaandhaynta iimaylka spam
    • Tilmaanta database-ka
  6. Moodooyinka fidinta casriga ah, fikradda cGAN ee qaboojinta badanaa waxay u muuqataa qaabkee?
    • Miisaanka aan tooska ahayn goynta
    • Feejignaanta ka gudubka qoraalka degdegga ah
    • Rinjiyeynta gacanta ee pixel kaliya
    • Deminta jadwalka buuqa
  7. Why must a conditional discriminator see both the generated image and its conditioning label?
    • It uses the label to choose the learning-rate schedule
    • It converts the label into a random noise seed only
    • It needs to judge whether the image is realistic and whether it matches the requested condition
    • It prevents the generator from receiving noise
  8. Besides concatenating a class label to the noise vector, which conditioning method is described in the guide?
    • A spectral-radius constraint
    • An inverse Hessian update
    • A sequence-level BLEU reward
    • A projection layer that scores compatibility between label and image features