Guide-yakabatana mibvunzo · Pakati Nhanho

StyleGAN Architecture Quiz

Ongorora StyleGAN mepu uye synthesis, layerwise maitiro musanganiswa, stochastic ruzha, uye shanduko yekuvaka muStyleGAN2 uye StyleGAN3.

Nzira dzekutungamira dzinoenderanaStylegan Architecture
Mubvunzo 1 ye 8

Chii chinangwa cheStyleGAN's mepu network?

Tarisa mubvunzo wese8 mibvunzo

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

  1. Chii chinangwa cheStyleGAN's mepu network?
    • Kuisa mumapoka kuti mufananidzo ndewechokwadi here kana kuti wemanyepo
    • Kusimudzira mifananidzo kune yakakwira resolution
    • Kumanikidza mufananidzo wekupedzisira wekuchengetedza
    • Kushandura iyo yakadzika z kuita imwe nzvimbo yakapatsanuka yepakati W
  2. Mune yekutanga StyleGAN, maitiro anopinzwa sei mumamepu emhando?
    • Nekubatanidza iyo latent vector kune yega pixel
    • Kuburikidza neAdaptive Instance Normalization (AdaIN)
    • Via yekupedzisira softmax layer
    • Kushandisa max pooling
  3. Ko iyo synthesis network inotanga kubva pachinzvimbo chelatent vector?
    • A akadzidza anogara ari tensor
    • Mufananidzo wechiso chaicho
    • Ruzha ruchena rwakachena saizi yakafanana neyakabuda
    • Mepu yemupendero
  4. Ko kugadzirisa masitayera payakaomarara (yakaderera-resolution) zvidimbu zvinonyanya kudzora?
    • Kugadzirwa kweganda rakanaka uye magwenzi
    • Ruvara rwemapikisi ega ega chete
    • Hunhu hwepamusoro-soro senge pose uye chimiro chechiso
    • Image file format
  5. Nderipi dambudziko rakagadziriswa neStyleGAN2 maererano nepakutanga?
    • Yakawedzera mavara-kune-mufananidzo kukurudzira
    • Yakabvisa hunhu hwakaita senge blob-yakagadzirwa neAdaIN
    • Kwakabvisa kudikanwa kwomunhu anosarura
    • Zvakaita kuti kudzidziswa kusatarisirwe zvizere kekutanga
  6. Chii chinonzi 'style mixing' muStyleGAN?
    • Kubatanidza ma dataset maviri usati wadzidziswa
    • Avhareji yezvakabuda zvemusaruri
    • Kusanganisa odhiyo nemufananidzo wakagadzirwa
    • Kushandisa akasiyana w vectors pamatanho akasiyana kubatanidza hunhu kubva pamifananidzo miviri
  7. What role do separate per-layer noise inputs play in original StyleGAN synthesis?
    • They replace the mapping network's entire style representation
    • They determine the file's compression format
    • They supply stochastic fine detail, such as varying hair strands or freckles
    • They provide the discriminator's real-image labels
  8. What visual behavior did StyleGAN3's alias-free design target?
    • Texture sticking and aliasing that make details move unnaturally as image content changes
    • The absence of a text tokenizer in original StyleGAN
    • A requirement that every output be a grayscale photograph
    • The need to store generated images as audio waveforms