Artificial Neural Networks for Vision ( Invited Paper )

For designing artificial visual systems that have higher functions like the human brain, it is important to understand the mechanism of the brain and learn from the real biological basis of it. Modeling neural networks plays an important role for this purpose. This paper introduces some results of this approach from our recent works: Improving the recognition rate of the neocognitron by the use of interpolating vectors, and Extraction of visual motion and optic flow.

[1]  D Marr,et al.  Directional selectivity and its use in early visual processing , 1981, Proceedings of the Royal Society of London. Series B. Biological Sciences.

[2]  Kunihiko Fukushima,et al.  Neocognitron: A hierarchical neural network capable of visual pattern recognition , 1988, Neural Networks.

[3]  Keiji Tanaka,et al.  Integration of direction signals of image motion in the superior temporal sulcus of the macaque monkey , 1986, The Journal of neuroscience : the official journal of the Society for Neuroscience.

[4]  Kunihiko Fukushima,et al.  Interpolating vectors for robust pattern recognition , 2007, Neural Networks.

[5]  Kunihiko Fukushima,et al.  Neocognitron for handwritten digit recognition , 2003, Neurocomputing.

[6]  R. Wurtz,et al.  Sensitivity of MST neurons to optic flow stimuli. II. Mechanisms of response selectivity revealed by small-field stimuli. , 1991, Journal of neurophysiology.

[7]  G. Orban,et al.  The spatial distribution of the antagonistic surround of MT/V5 neurons. , 1997, Cerebral cortex.

[8]  Kunihiko Fukushima,et al.  Neural network model for extracting optic flow , 2005, Neural Networks.

[9]  Kunihiko Fukushima,et al.  Neural Network Model for Extracting Optic Flow , 2005, ICANN.

[10]  Kunihiko Fukushima,et al.  Robust Pattern Recognition by Interpolating Vectors , 2007, 2007 International Joint Conference on Neural Networks.