BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:BMI Prize 2020 SEMINAR // Adrien Doerig "Crowding and the Architec
 ture of the Visual System"
DTSTART:20201202T121500
DTEND:20201202T131500
DTSTAMP:20260506T144859Z
UID:a452900cebd7bd7d1fe8690dbfd863c9bd20f3a9952fe5b79615ab1f
CATEGORIES:Conferences - Seminars
DESCRIPTION:Adrien Doerig\, Laboratory of Psychophysics\, BMI\, EPFL Host 
 : R. Schneggenberger\nClassically\, vision is seen as a cascade of local\,
  feedforward computations. This framework has been tremendously successful
 \, inspiring a wide range of ground-breaking findings in neuroscience and 
 computer vision. Recently\, feedforward Convolutional Neural Networks (ffC
 NNs)\, inspired by this classic framework\, have revolutionized computer v
 ision and been adopted as tools in neuroscience. However\, despite these s
 uccesses\, there is much more to vision. I will present our work using vis
 ual crowding and related psychophysical effects as probes into visual proc
 esses that go beyond the classic framework. In crowding\, perception of a 
 target deteriorates in clutter. We focus on global aspects of crowding\, i
 n which perception of a small target is strongly modulated by the global c
 onfiguration of elements across the visual field. We show that models base
 d on the classic framework\, including ffCNNs\, cannot explain these effec
 ts for principled reasons and identify recurrent grouping and segmentation
  as a key missing ingredient. Then\, we show that capsule networks\, a rec
 ent kind of deep learning architecture combining the power of ffCNNs with 
 recurrent grouping and segmentation\, naturally explain these effects. We 
 provide psychophysical evidence that humans indeed use a similar recurrent
  grouping and segmentation strategy in global crowding effects. In crowdin
 g\, visual elements interfere across space. To study how elements interfer
 e over time\, we use the Sequential Metacontrast psychophysical paradigm\,
  in which perception of visual elements depends on elements presented hund
 reds of milliseconds later. We psychophysically characterize the temporal 
 structure of this interference and propose a simple computational model. O
 ur results support the idea that perception is a discrete process. Togethe
 r\, the results presented here provide stepping-stones towards a fuller un
 derstanding of the visual system by suggesting architectural changes neede
 d for more human-like neural computations.\n 
LOCATION:Online https://epfl.zoom.us/meeting/register/tJIvceqhqjMpGtGHWZQ7
 gmu96_V8MXMdgpnX
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR