BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Self-supervised Learning from Images\, Videos\, and a single Image
  plus Augmentations - Dr. Yuki M. Asano (University of Amsterdam)
DTSTART:20220831T153000Z
DTEND:20220831T163000Z
UID:TALK178067@talks.cam.ac.uk
CONTACT:Gwangbin Bae
DESCRIPTION:*Abstract*\n\nIn this talk I will talk about pushing the limit
 s of what can be learnt without using any human annotations. After a first
  overview of what self-supervised learning is\, we will first dive into ho
 w clustering can be combined with representation learning using optimal tr
 ansport and analyize how semantic the resulting clusters are ([1] at ICLR'
 20). We will then talk about how multi-modal data can be leveraged for det
 ecting objects fully unsupervisedly. Finally\, as augmentations are crucia
 l for all of self-supervised learning\, we will analyze these in more deta
 il in recent preprint [3]. Here\, we show that it is possible to extrapola
 te to semantic classes such as those of ImageNet using just a single datum
  as visual input when combined with strong augmentations. \n\n[1] https://
 arxiv.org/abs/1911.05371\n\n[2] https://arxiv.org/abs/2104.06401\n\n[3] ht
 tps://arxiv.org/abs/2112.00725 \n\n*Bio*\n\nYuki Asano is an assistant pro
 fessor for computer vision and machine learning at the Qualcomm-UvA lab at
  the University of Amsterdam\, where he works with Cees Snoek\, Max Wellin
 g and Efstratios Gavves. His current research interests are multi-modal an
 d self-supervised learning and ethics in computer vision. Prior to his cur
 rent appointment\, he finished his PhD at the Visual Geometry Group (VGG) 
 at the University of Oxford working with Andrea Vedaldi and Christian Rupp
 recht. During his time as a PhD student he also interned at Facebook AI Re
 search and worked at TransferWise. Prior to the PhD he studied physics at 
 the University of Munich (LMU) and Economics in Hagen as well as a MSc in 
 Mathematical Modelling and Scientific Computing at the Mathematical Instit
 ute in Oxford.\n\n*Location*\n\nThe talk will be given at *Lecture Theatre
  1 (LT1)* in the Engineering Department (Trumpington St\, Cambridge CB2 1P
 Z).\n\n*Zoom link*\n\nZoom link: https://zoom.us/j/6492509351?pwd=U0hoSzJ0
 anlhRGhzYVFmTzltNk9wZz09\n\nMeeting ID: 649 250 9351 / Passcode: 7mu5ZJ\n\
 n*Google Calendar*\n\nTo get updates on future seminars\, please subscribe
  to the following Google calendar:\nhttps://calendar.google.com/calendar/u
 /0?cid=c2pjcHN0YXM2N3QyMWU3c2FqNjBqNWNiYXNAZ3JvdXAuY2FsZW5kYXIuZ29vZ2xlLmN
 vbQ\n
LOCATION:Lecture Theatre 1 (LT1)\, Engineering Department\, Trumpington St
 \, Cambridge (CB2 1PZ)
END:VEVENT
END:VCALENDAR