quiz bu lëkkale ak tegtal · Dëgër Niveau

Laajte ci xaaj panoptik

Xoolaat segmentaasioŋ panoptik ngir mbir ak yeneen mbir, misaal ID ak Kalite Panoptik; Jàngaleel ni ñaari mbir yi ci metric di wane njuumte yu wuute.

Yooni njiit yu jëm ciSegmentaasioŋ panoptik
Laaj 1 de 8

Lan la segmentaasioŋ panoptik di jox pixel bu nekk ci nataal bi?

Xoolaatal laaj bu nekk8 laaj

Sooy def quiz bi, tontu yi ak leeral yi dañuy feeñ.

  1. Lan la segmentaasioŋ panoptik di jox pixel bu nekk ci nataal bi?
    • Etiketu kategori, boole ci ID misaal ngir mbir yuñ mëna lim
    • xóotaayu valeur kese
    • Benn boyet buy tënk
    • Niveau de flou
  2. Ci terminoloji panoptik, lan mooy klaas 'mbiri àdduna'?
    • Mbir yuñ mëna lim lu melni oto ak nit
    • Gox yu amorphe yu melni asamaan, tali, wala ñax
    • Etiketu mbind yi ci biir nataal bi
    • Métadone yu kamera
  3. Lan mooy jafe-jafe bi gëna mag ci xaaj semantik bi xaaj panoptik di daan?
    • Mënul màndargaal asamaan
    • Ci wideo rek lay dox
    • Mënul wuutale misaal yi ci benn kategori
    • Dafay baña bayyi xel ci ginaaw
  4. Ban metrik lañuy jëfandikoo ngir jàngat kalite segmentaasioŋ panoptik?
    • Kadre ci segond bu nekk
    • Rasio kompresioŋ
    • Kulòor bu jaar yoon
    • Kalite panoptik (PQ)
  5. What does segmentation quality (SQ) measure within Panoptic Quality?
    • The average IoU of matched predicted and ground-truth segments
    • The number of images in the dataset
    • The speed of inference
    • The fraction of classes that are things
  6. Ban architecture bu bees mooy jëfandikoo mbir yi ak mbir yi ci benn model buñ boole muy lalu ci laaj?
    • JPEG
    • Mask2
    • ResNet-18 kese
    • Seggal bu sukkandiko ci sàrt
  7. How is Panoptic Quality (PQ) decomposed in the original formulation?
    • As image resolution multiplied by object count
    • As precision plus recall only
    • As segmentation quality multiplied by recognition quality
    • As frame rate divided by latency
  8. Why can a single PQ score conceal useful diagnostic information?
    • PQ measures only image brightness
    • PQ excludes stuff classes
    • PQ is not computed from predictions
    • Segmentation and recognition errors can contribute differently to the aggregate