Guide-yakabatana mibvunzo · Pakati Nhanho

ESRGAN uye GAN Super-Resolution Quiz

Dzidza ESRGAN's adversarial image-upscaling dhizaini, yayo yekufungidzira realism tradeoff, uye miganhu yezvakagadzirwa zvayo.

Nzira dzekutungamira dzinoenderanaEsrgan uye GAN Super Resolution
Mubvunzo 1 ye 8

Ko iyo 'GAN' muESRGAN inorevei?

Tarisa mubvunzo wese8 mibvunzo

Mhinduro uye tsananguro zvinoonekwa paunenge uchitora bvunzo.

  1. Ko iyo 'GAN' muESRGAN inorevei?
    • Generative Adversarial Network
    • Gradient Accumulation Network
    • Gaussian Approximation Node
    • Global Attention Network
  2. Ndechipi chivakwa chekuvaka chinoshandiswa ne ESRGAN jenareta?
    • Masero eMemory Yenguva Yakareba Yakareba
    • Residual-in-Residual Dense Blocks (RRDB)
    • Capsule routing layers
    • Transformer self-attention blocks
  3. Sei vanyori veESRGAN vakabvisa batch normalization kubva kune jenareta?
    • Zvakaderedza kudzidzira zvakanyanya
    • Yakawedzera ndangariro kupfuura miganhu
    • Zvaida kuunza zvinhu zvinoonekwa
    • Yakanga isingaenderane nemaGAN
  4. Chii chikuru chekutengesa-off ESRGAN chinoita zvichienzaniswa nepixel-kurasikirwa-chete nzira?
    • Speed pamusoro pekururama
    • Kurongeka kwemavara pamusoro pekugadziriswa
    • Yakaderera ndangariro pamusoro pepamusoro resolution
    • Perceptual realism pamusoro pechokwadi pixel kutendeka
  5. Ndekupi kurasikirwa kweESRGAN's perceptual (VGG) kuyerwa?
    • Pazvinhu zvisati zvaitika
    • Pamapikisheni eRGB mbishi chete
    • Pachibodzwa chekupedzisira chemusaruri
    • PaFourier-shanduko yekupinda
  6. Ndeupi hutsva hwakaunzwa nemusarura weESRGAN?
    • A chete convolutional autoencoder
    • A relativistic discriminator achienzanisa real vs. fake realism
    • Iyo yekusimbisa-yekudzidza mubairo modhi
    • A frozen pretrained classifier
  7. Which competition did the ESRGAN paper report winning?
    • The ImageNet classification challenge
    • The Vesuvius Challenge
    • The PIRM 2018 perceptual super-resolution challenge
    • The COCO image-captioning challenge
  8. Why should generated detail from ESRGAN not be treated as recovered photographic evidence?
    • The model records the original sensor data exactly
    • GANs can only enlarge images without changing pixels
    • The output is guaranteed to match the lost high-resolution image
    • The model can synthesize plausible textures that were not present in the low-resolution input