ESRGAN ati GAN Super-opinnu adanwo

Awọn idahun ati awọn alaye han bi o ṣe n mu ibeere naa.

1. Kini 'GAN' ni ESRGAN tọka si?
   • Generative Adversarial Network
   • Nẹtiwọọki ikojọpọ Gradient
   • Gaussian isunmọ Node
   • Agbaye akiyesi Network
2. Ohun amorindun ile wo ni olupilẹṣẹ ESRGAN nlo ni akọkọ?
   • Awọn sẹẹli Iranti Igba Kukuru gigun
   • Awọn bulọọki iponju ti o ku ninu-iku (RRDB)
   • Awọn fẹlẹfẹlẹ ipa ọna Capsule
   • Awọn bulọọki akiyesi ara ẹni Amunawa
3. Kini idi ti awọn onkọwe ESRGAN ṣe yọkuro deede ipele lati inu olupilẹṣẹ naa?
   • O fa fifalẹ ikẹkọ pupọ
   • O pọ si iranti kọja awọn opin
   • O nifẹ lati ṣafihan awọn ohun-ọṣọ wiwo
   • O ko ni ibamu pẹlu awọn GAN
4. Kini iṣowo akọkọ-pipa ESRGAN ṣe akawe si awọn ọna piksẹli-pipadanu-nikan?
   • Iyara lori išedede
   • Awọ išedede lori ipinnu
   • Isalẹ iranti lori ti o ga
   • Otitọ oye lori iṣotitọ ẹbun gangan
5. Nibo ni a ti ṣe iwọn pipadanu ESRGAN's perceptual (VGG)?
   • Lori awọn ẹya ṣaaju ṣiṣe
   • Lori awọn piksẹli RGB aise nikan
   • Lori awọn discriminator ká ik Dimegilio
   • Lori Fourier-iyipada awọn igbewọle
6. Atunse wo ni iyasoto ti ESRGAN ṣafihan?
   • A odasaka convolutional autoencoder
   • Iyatọ ti o ni ibatan ti o ṣe afiwe gidi la
   • Awoṣe ẹ̀kọ́ ìmúlẹ̀mọ́
   • A tutunini classifier pretrained
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