Maswali yanayounganishwa na mwongozo · Kati Kiwango

Maswali ya Kupoteza Malengo kwa Maswali ya Utambuzi Isiyosawazishwa

Hukagua kibadilishaji cha focal-hasara, usawa wa darasa, jukumu la alfa na gamma, na jinsi mifano iliyoainishwa vizuri na ngumu hutofautiana.

Njia za mwongozo zinazohusianaUpotezaji wa Kuzingatia Kwa Utambuzi Usio na Usawazishaji
Swali 1 ya 8

Ni tatizo gani la msingi katika ugunduzi wa kitu cha hatua moja ambalo upotezaji wa msingi uliundwa kushughulikia?

Hakiki kila swali8 maswali

Majibu na maelezo huonekana unapojibu maswali.

  1. Ni tatizo gani la msingi katika ugunduzi wa kitu cha hatua moja ambalo upotezaji wa msingi uliundwa kushughulikia?
    • Usawazishaji uliokithiri kati ya vitu adimu vya mbele na mandharinyuma tele
    • Kasi ya polepole ya uelekezaji wa vigunduzi vya hatua mbili
    • Ukosefu wa data ya mafunzo yenye lebo
    • Kuzidisha kwa vitu vikubwa
  2. Ni kigezo gani cha kurekebisha ambacho upotezaji wa kielelezo huzidisha kwenye mtambuka?
    • logi(p_t)
    • (1 - p_t)^gamma
    • p_t^2 pekee
    • 1 / p_t
  3. Ni nini hufanyika kwa upotezaji wa umakini wakati kigezo cha kulenga gamma kimewekwa kuwa 0?
    • Inakuwa kosa la maana la mraba
    • Inapuuza mifano yote chanya
    • Inakuwa kawaida msalaba-entropy
    • Inatofautiana hadi isiyo na mwisho
  4. Kuongeza gamma kunafanya nini kwa mifano rahisi, iliyoainishwa vyema?
    • Huongeza mchango wao wa hasara
    • Huacha mchango wao bila kubadilika
    • Huziondoa kwenye mkusanyiko wa data
    • Inakandamiza mchango wao wa hasara kwa nguvu zaidi
  5. Ni usanifu gani wa kigunduzi ulianzisha upotezaji wa msingi?
    • RetinaNet
    • YOLOv1
    • Kasi ya R-CNN
    • AlexNet
  6. Je, parameta ya alpha huwa na jukumu gani pamoja na gamma?
    • Inaweka kiwango cha kujifunza
    • Inasawazisha uzito wa jamaa wa madarasa chanya dhidi ya hasi
    • Inadhibiti azimio la picha
    • Inafafanua idadi ya masanduku ya nanga
  7. For a hard example with p_t close to zero, what is the focal modulating factor (1 − p_t)^gamma likely to do?
    • Drive the factor to zero
    • Turn the example into a positive label
    • Keep the factor relatively close to one so the example retains substantial loss weight
    • Replace the cross-entropy term with alpha
  8. A dense detector has many easy background locations. What effect does focal loss aim to have on their combined training influence?
    • Increase their count by duplicating them
    • Remove all positive object examples
    • Make every example receive an identical loss
    • Reduce their dominance so training can focus more on sparse hard examples