Downloads
Amodal completion on border-ownership surround model
This package of Matlab code contains the model of border-ownership selective cells, the original model of which was published in [1]. A more realistic model that is capable of determining BO from natural image patches was reported in [2]. A dynamic model with spiking neurons was published in [3]. Please cite these two papers when you refer to this model and/or package. The present package contains the experimental settings and stimuli that enable the reproduction of electrophysiological experiment reported in [4].
The model determines the direction of BO with respect to a contour based on surround modulation. The package provides 300 surround regions with randomly-chosen Gaussians. The package also includes stimuli for amodal completion.
All codes are copyrighted. One may download, run and modify the codes for academic purposes only. Please cite the references [1] and [2] when you publish/present the work that used this package. Please note that we do not take any responsibility of any incident that may be caused by the code, including bugs in the code.
[1] Surrounding suppression and facilitation in the determination of border ownership. K. Sakai and H. Nishimura (2006) Journal of Cognitive Neuroscience, Vol.18, No.4, 562-579 [2] Consistent and robust determination of border-ownership based on asymmetric surrounding modulation. K. Sakai, H. Nishimura, R. Shimizu and K. Kondo (2012) Neural Networks, Vol.33, 257-274 [3] Latency Modulation of Border Ownership Selective Cells in V1-V2 Feed-Forward Model. K. Sakai and S. Michii (2013) Lecture Notes in Computer Science (ICONIP 2013), Vol.8228, pp.291-300, doi: 10.1007/978-3-642-42051-1_37 [4] Analysis of the context integration mechanisms underlying figure-ground organization in visual cortex. N. R. Zhang and R. von der Heydt (2010) Journal of Neuroscience, Vol.30, No.19, 6482-6496
ソースコード
モデルのソースコードはVisiomeにアップロードしてあります。