Idanwo ti o sopọ mọ itọsọna · Alabọde Ipele

Ti kii-O pọju Bomole Quiz

Review greedy NMS, Intersection over Union, confidence ordering, crowded-scene threshold tradeoffs, and soft-score suppression.

Jẹmọ itọsọna awọn ọnaTi kii ṣe Imukuro ti o pọju
Ibeere 1 ti 8

Kini idi akọkọ ti Idasilẹ ti kii-O pọju ni wiwa ohun?

Awotẹlẹ gbogbo ibeere8 ibeere

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

  1. Kini idi akọkọ ti Idasilẹ ti kii-O pọju ni wiwa ohun?
    • Lati yọ awọn apoti agbekọja laiṣe, titọju ọkan fun ohun kan
    • Lati mu awọn nọmba ti ri apoti
    • Lati tunṣe iwọn aworan titẹ sii
    • Lati kọ awọn iwuwo aṣawari
  2. Metiriki wo ni NMS boṣewa lo lati pinnu boya awọn apoti meji ni lqkan pupọ ju?
    • Euclidean ijinna laarin awọn ile-iṣẹ
    • Ikorita lori Union (IoU)
    • Cosine ibajọra ti awọn ẹya ara ẹrọ
    • Iyatọ awọ Pixel
  3. Ni NMS ojukokoro, apoti wo ni a yan ni akọkọ ni yika kọọkan?
    • Apoti ti o tobi julọ nipasẹ agbegbe
    • Apoti ti o sunmọ si aarin aworan
    • Apoti pẹlu Dimegilio igbẹkẹle ti o ga julọ
    • A laileto yàn apoti
  4. Kini ailera ti a mọ ti NMS greedy ni awọn oju iṣẹlẹ ti o kunju?
    • Ko le mu awọn aworan awọ mu
    • O nigbagbogbo fun wa ju diẹ apoti lapapọ
    • O nilo ikẹkọ GPU lati ṣiṣẹ
    • O le pa ohun gidi kan nitosi ti o bori apoti ti o tọju
  5. Bawo ni Soft-NMS ṣe yatọ si NMS boṣewa?
    • O bajẹ awọn ikun ti awọn apoti agbekọja dipo yiyọ wọn kuro patapata
    • O npa gbogbo awọn apoti laiwo ti Dimegilio
    • O foju awọn ikun igbekele patapata
    • O ṣiṣẹ nikan lakoko ikẹkọ
  6. Kini idi ti NMS ni igbagbogbo lo lọtọ fun kilasi ohun kọọkan?
    • Nitoripe kilasi kọọkan nilo ipinnu aworan ti o yatọ
    • Nitorinaa apoti ti kilasi kan ko dinku apoti ti kilasi ti o yatọ
    • Lati ṣe algorithm losokepupo lori idi
    • Nitori IoU ko le ṣe iṣiro kọja awọn kilasi
  7. How is Intersection over Union (IoU) calculated for two detection boxes?
    • Union area divided by the intersection area
    • The sum of both box perimeters
    • Intersection area divided by the union area
    • The distance between the box centers
  8. What is a typical effect of lowering the IoU threshold in greedy NMS?
    • Fewer boxes are suppressed in all cases
    • The detector retrains its class labels
    • The confidence scores are converted into image pixels
    • More overlapping boxes are suppressed, which can risk removing nearby real objects