BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:MechE Colloquium: Opponency Revisited
DTSTART:20201027T121500
DTEND:20201027T131500
DTSTAMP:20260413T103227Z
UID:cb5ad7573233dc3b5bcbd2d7ce68ee64d1005033f006e29f39e58fac
CATEGORIES:Conferences - Seminars
DESCRIPTION:Prof. Sabine Süsstrunk\, Image and Visual Representation Labo
 ratory\, EPFL School of Computer and Communication Sciences (IC)\, Institu
 te of Computer and Communication Sciences (IINFCOM)\nAbstract:\nAccording 
 to the efficient coding hypothesis\, the goal of the visual system should 
 be to encode the information presented to the retina with as little redund
 ancy as possible. From a signal processing point of view\, the first step 
 in removing redundancy is de-correlation\, which removes the second order 
 dependencies in the signal. This principle was explored in the context of 
 trichromatic vision by Buchsbaum and Gottschalk (1) and later Ruderman et 
 al. (2) who found that linear de-correlation of the LMS cone responses mat
 ches the opponent color coding in the human visual system.\n\nAnd yet\, th
 ere is comparatively little research in image processing and computer visi
 on that explicitly model and incorporate color opponency into solving imag
 ing tasks. A common perception is that “colors” are redundant and/or t
 oo correlated to be of any interest\, or that they are too complex to deal
  with. Within deep learning frameworks\, color features are rarely conside
 red.\n\nIn this talk\, I will illustrate with several simple algorithms\, 
 such as saliency and super-pixels\, that considering opponent colors can s
 ignificantly improve image processing and computer vision tasks not only i
 n image enhancement but also image segmentation\, image ranking\, etc. We 
 have in addition extended the concept of “color opponency” to include 
 near-infrared. And we found that de-correlation concepts also apply to dee
 p learning models in rather interesting ways.\n\n(1) http://rspb.royalsoci
 etypublishing.org/content/220/1218/89.short\n(2) http://www.opticsinfobase
 .org/josaa/abstract.cfm?&uri=josaa-15-8-2036\n\n\nBio:\nSabine Süsstrunk 
 is a full professor and leads the Image and Visual Representation Lab in t
 he School of Computer and Communication Sciences\n(IC) at EPFL since 1999.
  From 2015-2020\, she was also Director of the Digital Humanities Institut
 e (DHI)\, College of Humanities (CdH). Her main research areas are in comp
 utational photography\, computational imaging\, color image processing and
  computer vision\, machine learning\, and computational image quality and 
 aesthetics. Sabine has authored and co-authored over 150 publications\, of
  which 7 have received best paper/demo awards\, and holds over 10 patents.
 \n\nSabine served as chair and/or committee member in many international c
 onferences on image processing\, computer vision\, and image systems engin
 eering. She is Founding Member and Member of the Board (President 2014-201
 8) of the EPFL-WISH (Women in Science and Humanities) Foundation\, Member 
 of the Foundation Council of the SNSF (Swiss National Science Foundation)\
 , Member of the Board of the SRG SSR (Swiss Radio and Television)\, and Me
 mber of the Board of Largo Films. She received the IS&T/SPIE 2013 Electron
 ic Imaging Scientist of the Year Award for her contributions to color imag
 ing\, computational photography\, and image quality\, and the 2018 IS&T Ra
 ymond C. Bowman Award for dedication in preparing the next generation of i
 maging scientists. Sabine is a Fellow of IEEE and IS&T.
LOCATION:Zoom webinar https://epfl.zoom.us/s/84160054600
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR