Laaj ak tontu ci nataal

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

1. Lan la sistemu sous-titre nataal di defar ci génne gi?
   • Tegtal ci làkk wiñ nàmp ci nataal bi
   • Benn xeetu nataal bu gëna rëy
   • Limu coordonnée pixel yi kese
   • Fichier buñ komprime ci nataal bi
2. Ci biir pipeline biy bind sottite, ban cër la encodeur visuel bi di def?
   • Dafay bind frase bu mujj bi baat par baat
   • Dafay dindi màndarga yi ci nataal bi
   • Dafay tekki mbind yi ci diggante làkk yi
   • Dafay joxe poñ ci nàfar bi
3. Lan moo waral bàyyi xel ci nataal yi am njariñ?
   • Dafay kompresse nataal bi ngir denc
   • Dina dindi bépp soxla ci done tàggat yaram
   • Dafay may model bi mu sëgg ci gox nataal yi ci baat bu nekk
   • Dafay soppi kàddu yi ci audio
4. Ban setu done lañu gëna jëfandikoo ngir tàggat ak jàngat mbindu nataal yi?
   • MNIST
   • Wikitekst
   • Kaddu Libri
   • MS COCO
5. Lan mooy tekki dekodeeru kapsiyoŋ ngir defar ay baat 'ci anam wu dellu ginaaw'?
   • Dafay wax luy waaja am ci baat bu nekk ci jëfandikoo nataal bi ak baat yiñ defar ba leegi
   • Dafay defar kàddu yépp benn yoon
   • Benn baat kese lay génne ci nataal bu nekk
   • Dafay delloosi nataal bi ci pixel yi
6. Benn anam bu ñuy faral lijjanti ci xeetu mbind yi ñuy woowe 'hallucination' dafay wax ci lan?
   • Defar ay mbind yu gàtt lool
   • Tegtal mbir yu nekkul ci nataal bi
   • Tekki mbind yi ci anam wu jaarul yoon
   • Dawal ndànk lool ci nataal yu yaatu
7. How many human captions were collected for each image in the original COCO Captions dataset?
   • One caption total for the entire dataset
   • No text descriptions were collected
   • Five captions per image
   • One hundred captions for every image
8. What does CIDEr compare when evaluating an image caption?
   • Only the image’s pixel resolution
   • The caption’s audio waveform against a transcript
   • Whether the caption contains a certain number of words
   • The candidate caption against human reference captions using consensus-weighted n-grams