Abstract
We have reported previously that the performance of a neocognitron can be improved by a built-in bend-extracting layer. The conventional bend-extracting layer can detect bend points and end points of lines correctly, but not always crossing points of lines. This paper shows that an introduction of a mechanism of disinhibition can make the bend-extracting layer detect not only bend points and end points, but also crossing points of lines correctly. This paper also demonstrates that a neocognitron with this improved bend-extracting layer can recognise handwritten digits in the real world with a recognition rate of about 98%. We use the technique of dual thresholds for feature-extracting S-cells, and higher threshold values are used in the learning than in the recognition phase. We discuss how the threshold values affect the recognition rate.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Explore related subjects
Discover the latest articles, news and stories from top researchers in related subjects.References
Fukushima K. Neocognitron: a hierarchical neural network capable of visual pattern recognition. Neural Networks 1988; 1: 119–130.
Fukushima K, Miyake S. Neocognitron: a new algorithm for pattern recognition tolerant of deformations and shifts in position. Pattern Recognition 1982; 15: 455–469.
Hubel DH, Wiesel TN. Receptive fields and functional architecture in nonstriate areas (18 and 19) of the cat. J Neurophysiology 1965; 28: 229–289.
Fukushima K, Wake N. Improved neocognitron with bend-detecting cells. Int Joint Conf Neural Networks (IJCNN'92), Baltimore, vol IV. 1992; 190–195.
Fukushima K. Analysis of the process of visual pattern recognition by the neocognitron. Neural Networks 1989; 2: 413–420.
Fukushima K, Tanigawa M. Use of different thresholds in learning and recognition. Neurocomputing 1996; 11: 1–17.
Fukushima K, Nagahara K, Shouno H, Okada M. Training neocognitron to recognise handwritten digits in the real world. World Congress on Neural Networks (WCNN'96), San Diego, 1996; 21–24.
Fukushima K, Nagahara K, Shouno H. Training neocognitron to recognise handwritten digits in the real world. Proc Second Aizu Int Symposium on Parallel Algorithms/Architectures Synthesis (pAs'97), Aizu-Wakamatsu, Japan, 1997; 292–298.
LeCun Y, Boser B, Denker JS, Henderson D, Howard RE, Hubbard W, Jackel LD. Backpropagation applied to handwritten zip code recognition. Neural Computation 1989; 1: 541–551.
Fukushima K, Wake N. Handwritten alphanumeric character recognition by the neocognitron. IEEE Trans Neural Networks 1991; 2: 355–365.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Fukushima, K., Kimura, E. & Shouno, H. Neocognitron with improved bend-extractors: Recognition of handwritten digits in the real world. Neural Comput & Applic 7, 260–272 (1998). https://doi.org/10.1007/BF01414887
Issue Date:
DOI: https://doi.org/10.1007/BF01414887