A neural model for detection of temporal structure. (CROSBI ID 560443)
Prilog sa skupa u zborniku | izvorni znanstveni rad | međunarodna recenzija
Podaci o odgovornosti
Domijan, Dražen
engleski
A neural model for detection of temporal structure.
According to the temporal correlation hypothesis, synchronization of neural activity in different spatial maps solves the feature binding problem. Here, a new model of visual working memory is proposed which is able to group synchronous temporal events without neural synchronization. Instead, the model is based on the difference in the firing rate. The model integrates discrete inputs over time and compares activity in different integrators. When the amplitude difference in integrators is large enough due to the different rates of evidence accumulation, temporal figure and background are distinguished in the working memory. Computer simulations showed that the model correctly groups events according to their deterministic or stochastic temporal structure. The model is robust with respect to the temporal noise and to the correlation between figure-ground events. Also, the model is able to explain visual prior entry and perceptual asynchrony between colour, motion and orientation.
Temporal Integration; Neural Network; Computer Simulation; Temporal Structure
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
nije evidentirano
Podaci o prilogu
123-128.
2009.
objavljeno
Podaci o matičnoj publikaciji
Proceedings of the 25th Annual Meeting of the International Society for Psychophysics
Elliott, Mark A. ; Antonijević, Stanislava ; Berthaud, Sarah ; Mulcahy, Paul ; Martyn, Caitriona ; Bargery, Brian ; Schmidt, Heike
Galway: National University of Galway
Podaci o skupu
25th Annual Meeting of the International Society for Psychophysics
predavanje
21.10.2009-24.10.2009
Gaillimh, Irska