Presentation 2015-03-16
A simple visual model for optical-flow estimation generates illusory patterns
Daiki Nakamura, Shunji Satoh,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Computational models of vision should not only be able to reproduce the Input-Output relationship of the human vision or neural systems, but also should those models be able to predict the I/O of vision. First, we examine if an MT model proposed by authors can reproduce the dependency of background luminance of rotating drift-illusory patterns. Moreover, to evaluate the plausibility of the MT model, we generate an enormous set of possible visual patterns; 8^8=16,777,216 patterns, as inputs to the MT model. The numerical quantities of model outputs by computer simulation for 8^8 inputs can be used as estimation of humans' perception. By psychophysical experiments, we show that the model prediction is consistent with human perception.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Visual illusion / Lucas-Kanade method / MT / Computational model
Paper # MBE2014-143,NC2014-94
Date of Issue

Conference Information
Committee NC
Conference Date 2015/3/9(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Neurocomputing (NC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A simple visual model for optical-flow estimation generates illusory patterns
Sub Title (in English)
Keyword(1) Visual illusion
Keyword(2) Lucas-Kanade method
Keyword(3) MT
Keyword(4) Computational model
1st Author's Name Daiki Nakamura
1st Author's Affiliation Graduate School of Information Systems, The University of Electro-Communications()
2nd Author's Name Shunji Satoh
2nd Author's Affiliation Graduate School of Information Systems, The University of Electro-Communications
Date 2015-03-16
Paper # MBE2014-143,NC2014-94
Volume (vol) vol.114
Number (no) 515
Page pp.pp.-
#Pages 6
Date of Issue