Quiz buy tontu ci laaj ak tontu

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

1. Yan ñaari mbir la sistemu tontu laaj yi ñuy gis di jël?
   • Nataal ak laaj ci làkk wiñ nàmp
   • Ñaari nataal yuñ wara méngale
   • Laaj ak clip audio
   • Benn nataal kese bu amul mbind
2. Lu tax ñu sos VQA v2.0 dataset ak tontu yu 'équilibré'?
   • Ngir defar nataal yu gëna am dayo
   • Ngir wàññi bët bu wuute ci làkk, model yi duñu mëna xalaat ci bind kese
   • Ngir dindi bépp laaj waaw/déedet
   • Ngir wàññi done yi ngir gëna gaaw ci tàggat yaram
3. Luy 'njaaxaanaay ci làkk' ci VQA?
   • Tekki laaj yi ci làkk wu jaarul yoon
   • Jëfandikoo làkk yu bari benn yoon
   • Tontu ci laaj ak tontu te doo bàyyi xel ci nataal bi
   • Dama taamu nataal yu gàtt yi moo gën yu gudd yi
4. naka la xeetu VQA yu bari di boolee nataal ak leerali laaj?
   • Ni ñuy boole pixel ak araf yu ñor
   • Soo defaree nataal bu bees ci laaj bi
   • Soo bañee bàyyi xel ci laaj bi
   • Ni ñuy boole seeni misaal yuñ kode, lu bari ci cross-attention
5. Sistem VQA yu yàgg yi dañu daan faral di joxe tontu ci def lan?
   • Xayma ci kaw ab vocabulaire bu fiks bu am tontu yuñ gëna xam
   • Desine ab boîte de limitation kese
   • Soppi nataalu njàngat mi
   • Delloo laaj biñ dugal te soppikuwul
6. Lan mooy misaalu jafe-jafe VQA bu gëna xarañ?
   • Soppi yaatuwaayu nataal ci nataal bu ndaw
   • Tontu laaj yi jëm ci mbind ak nimero yi ci biir nataalu këyit
   • Tànnal nataal yi ci alfabe ci turu fichier bi
   • Komprime fichier wideo
7. How does the original VQA benchmark score an answer when people phrase the same answer differently?
   • It requires an exact match to one fixed phrase only
   • It scores image resolution instead of answers
   • It compares the prediction with multiple human answers using a consensus-based score
   • It uses the model’s confidence as the ground truth
8. Why can high VQA accuracy still fail to prove visual grounding?
   • Images cannot be represented by neural networks
   • Questions are always randomly generated
   • The benchmark contains no answer labels
   • A model may exploit dataset or language regularities instead of the image