BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Predicting visible flicker in temporally changing images - Gyorgy 
 Denes\, University of Cambridge
DTSTART:20200123T141500Z
DTEND:20200123T150000Z
UID:TALK137587@talks.cam.ac.uk
CONTACT:65221
DESCRIPTION:Novel display algorithms such as low-persistence displays\, bl
 ack frame insertion\, and temporal resolution multiplexing introduce tempo
 ral change into images at 40-180 Hz\, on the boundary of the temporal inte
 gration of the visual system. This can lead to flicker\, a highly-objectio
 nable artefact known to induce viewer discomfort. The critical flicker fre
 quency (CFF) alone does not model this phenomenon well\, as flicker sensit
 ivity varies with contrast\, and spatial frequency\; a content-aware model
  is required. In this paper\, we introduce a visual model for predicting f
 licker visibility in temporally changing images. The model performs a mult
 i-scale analysis on the difference between consecutive frames\, normalizin
 g values with the spatio-temporal contrast sensitivity function as approxi
 mated by the pyramid of visibility. The output of the model is a 2D detect
 ion probability map. We ran a subjective flicker marking experiment to fit
  the model parameters\, then analyse the difference between two display al
 gorithms\, black frame insertion and temporal resolution multiplexing\, to
  demonstrate the application of our model.\n
LOCATION:SS03 Meeting Room\, Computer Laboratory
END:VEVENT
END:VCALENDAR