Ikibazo gihuza ibibazo · Hagati Urwego

Byimbitse Bitandukanijwe Imyanzuro Ikibazo

Koresha uburyo bwimbitse kandi butandukanye, gereranya ibiciro byo kwemeza, ubare urugero ruto, kandi umenye igishushanyo mbonera cyimikorere.

Inzira zijyanye no kuyoboraByimbitse Bitandukanijwe
Ikibazo 1 Bya 8

Ni izihe ntambwe ebyiri gutandukana gutandukanijwe byimbitse bigabanya imyanzuro isanzwe?

Reba buri kibazo8 ibibazo

Ibisubizo nibisobanuro bigaragara mugihe ufata ikibazo.

  1. Ni izihe ntambwe ebyiri gutandukana gutandukanijwe byimbitse bigabanya imyanzuro isanzwe?
    • Intambwe yo guhuriza hamwe ikurikirwa nintambwe isanzwe
    • Babiri bakurikiranye 3x3 hamwe na ReLU hagati yabo
    • Ubujyakuzimu bwimbitse muyunguruzi kuri buri muyoboro, hanyuma umuyoboro wa 1x1 uhuza
    • Kode ya encoder hamwe na decoder deconvolution
  2. Muntambwe yimbitse, inzira zinjira zifatwa gute?
    • Buri muyoboro uyungurura wigenga hamwe na kernel yayo
    • Imiyoboro yose yakusanyirijwe hamwe mbere
    • Imiyoboro ihindagurika mbere yo kuyungurura
    • Gusa umuyoboro wambere uremejwe
  3. Ku kanseri kenshi, ni iki gupfundikira umubare w’operasiyo gishobora gukurwaho n’igitabo cyerekana ku convolution ya 3×3 depthwise separable?
    • Inshuro zigera kuri 2 munsi
    • Inshuro 100 munsi
    • Ikoresha kubara byinshi, ntabwo ari munsi
    • Inshuro zigera kuri 8 kugeza kuri 9 munsi
  4. Ni uruhe ruhare rwo kwemeza ingingo (1x1) muri iki gishushanyo?
    • Kugabanya imiterere yikibanza
    • Guhuza amakuru kumiyoboro kuri buri mwanya
    • Ongeraho kodegisi ya posisiyo
    • Gukuraho amagambo abogamye
  5. Nubuhe bwubatsi buzwi bwamamaye bwimbitse gutandukana kubikoresho bigendanwa?
    • MobileNet
    • ResNet
    • AlexNet
    • LSTM
  6. An implementation applies only the depthwise spatial filters and omits the pointwise layer. Which capability is missing from that block?
    • Spatial filtering within each channel
    • Separate kernels for the input channels
    • Learned combination of information across channels
    • Local spatial feature extraction
  7. Using the guide’s formulas with M = 4 input channels and N = 16 outputs, how many multiply-adds per location do standard and separable 3×3 convolutions require?
    • Standard 100; separable 576
    • Standard 576; separable 100
    • Standard 576; separable 36
    • Standard 64; separable 100
  8. In the guide’s cost ratio 1/N + 1/9 for a 3×3 separable convolution, what happens as the output-channel count N becomes very large?
    • The ratio approaches 9
    • The ratio approaches zero
    • The ratio becomes independent of kernel size for every kernel
    • The ratio approaches 1/9