BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:The role of meta-learning for few-shot classification - Eleni Tria
 ntafillou\, Google Brain
DTSTART:20220629T100000Z
DTEND:20220629T113000Z
UID:TALK175505@talks.cam.ac.uk
CONTACT:Elre Oldewage
DESCRIPTION:While deep learning has driven impressive progress\, one of th
 e toughest remaining challenges is generalization beyond the training dist
 ribution. Few-shot learning is an area of research that aims to address th
 is\, by striving to build models that can learn new concepts rapidly in a 
 more "human-like" way. While many influential few-shot learning  methods w
 ere based on meta-learning\, recently progress has been made by simpler tr
 ansfer learning algorithms\, and it has been suggested in fact that few-sh
 ot learning might be an emergent property of large-scale models. In this t
 alk\, I will give an overview of the evolution of few-shot learning method
 s and benchmarks\, with an emphasis on the role of meta-learning on few-sh
 ot classification. I will discuss lessons learned from using larger and mo
 re diverse benchmarks for evaluation and trade-offs between different appr
 oaches\, closing with an open discussion about remaining challenges.
LOCATION:Cambridge University Engineering Department\, CBL Seminar room BE
 4-38
END:VEVENT
END:VCALENDAR