BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Towards Multimodal Technologies for the World
DTSTART:20230203T140000
DTEND:20230203T150000
DTSTAMP:20260407T051705Z
UID:7fb8efc29422f34ac2c1e1f1a7859ca5a79eb044c4cbe678b8a1a19a
CATEGORIES:Conferences - Seminars
DESCRIPTION:Emanuele Bugliarello\nThere has been an explosive growth of vi
 sion-and-language architectures in the last few years\, which are usually 
 trained on English captions paired with images from North America or Weste
 rn Europe.\nIn this talk\, Emanuele will first introduce a new protocol to
  collect culturally relevant images and captions\, which resulted in MaRVL
 \, a multimodal reasoning dataset in five diverse languages. He will then 
 discuss limitations of state-of-the-art models when evaluated on multiling
 ual data\, made possible by the IGLUE benchmark.\nFinally\, he will show t
 hat we can substantially improve zero-shot cross-lingual transfer by compr
 omising our ideals of multilingual multimodal data.\n\nEmanuele Bugliarell
 o received his MSc from the IC School at EPFL in 2018\, and he iscurrently
  a final-year PhD Fellow in the NLP Section at the University of Copenhage
 n. His research lies at the intersection of language and vision\, with a p
 articular interest in building models and creating resources that\nreprese
 nt the diversity of cultural and linguistic backgrounds.\n\n 
LOCATION:BC 229 https://plan.epfl.ch/?room==BC%20229
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR