Ikibazo gihuza ibibazo · Hagati Urwego

Ikibazo cya Fréchet Intangiriro Ikibazo

Ongera usuzume ibiranga FID-Umwanya wa Gaussiya hamwe nubunini bwa sample-size sensitivite, wongeyeho ubundi buryo bwo gusuzuma ibyakozwe-ishusho ubudahemuka no gutandukana.

Inzira zijyanye no kuyoboraIntera Intangiriro Intera
Ikibazo 1 Bya 8

Amanota yo hasi ya FID yerekana iki?

Reba buri kibazo8 ibibazo

Ibisubizo nibisobanuro bigaragara mugihe ufata ikibazo.

  1. Amanota yo hasi ya FID yerekana iki?
    • Imibare yakozwe mumibare yegereye iyishusho nyayo
    • Amashusho yakozwe ntabwo ari ibintu bifatika
    • Icyitegererezo cyatojwe ibihe bike
    • Amashusho afite imiterere ihanitse
  2. Nuwuhe muyoboro wateganijwe FID ikoresha mugukuramo ibiranga amashusho?
    • ResNet-50
    • Intangiriro-v3
    • VGG-16
    • BERT
  3. Nigute FID ivuga muri make buri shusho yibiranga amashusho mbere yo kubigereranya?
    • Nka impuzandengo imwe ya pigiseli agaciro
    • Nka Urutonde Urutonde rwamazina ya dosiye
    • Nka Gaussian itandukanye hamwe na vector hamwe na materix ya covariance
    • Nka histogramu yimiyoboro yamabara gusa
  4. Ni ubuhe butumwa bukomeye bw'amanota yatangiriye FID yakemuye?
    • Intangiriro amanota yatinze kubara
    • Intangiriro amanota yakoraga gusa kumyandiko
    • Intangiriro Amanota asabwa nta rezo ya neural
    • Intangiriro amanota ntabwo yigeze agereranya amashusho yakozwe namakuru yukuri
  5. When a generator produces less varied Inception features than the real images, how can that affect FID?
    • A covariance mismatch can contribute to a higher FID
    • FID must become zero
    • FID ignores feature diversity
    • FID automatically identifies the exact cause of the collapse
  6. In the Diffusion Models Beat GANs comparison, how was FID evaluated for ImageNet?
    • With 500 generated samples
    • With 50,000 generated samples
    • With 5,000 generated samples
    • With 5 million generated samples
  7. Why can a small evaluation set make an FID estimate unreliable?
    • FID requires every image to have a human label
    • The Inception network stops extracting features on small sets
    • Finite-sample bias can distort the estimated distance between feature distributions
    • Small sets force all covariance values to zero
  8. What does Kernel Inception Distance (KID) offer as an alternative in the guide?
    • A method for training the Inception-v3 network
    • A way to generate image captions
    • A measure of video frame rate
    • A metric used for evaluation when sample counts are smaller