Abstract
We propose a new convolutional neural network – the FTNet and explain its theoretical background referring to the theory of a higher degree F-transform. The FTNet is parametrized by kernel sizes, on/off activation of weights learning, the choice of strides or pooling, etc. It is trained on the database MNIST and tested on handwritten inputs. The obtained results demonstrate that the FTNet has better recognition accuracy than the automatically trained LENET-5. We have also analyzed the FTNet and LENET-5 rotation invariance.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsNotes
- 1.
Weighted average in case of convolutional layers. Weights are being learned.
- 2.
- 3.
- 4.
- 5.
The text of this and the following subsection is a free version of a certain part of [11] where the theory of a higher degree F-transform was introduced.
- 6.
The size is determined by only five non-zero values of \(A^{tr}\) on D.
- 7.
60000 training and 10000 testing.
References
Schmidhuber, J.: Neural Netw. 61, 85 (2015)
Krizhevsky, A., Sutskever, I., Hinton, G.E.: Imagenet classification with deep convolutional neural networks, In: Pereira, F., Burges, C.J.C., Bottou, L., Weinberger, K.Q. (eds.) Advances in Neural Information Processing Systems 25, pp. 1097–1105. Curran Associates, Inc. (2012)
Szegedy, C., Ioffe, S., Vanhoucke, V.: CoRR abs/1602.07261 (2016)
He, K., Zhang, X., Ren, S., Sun, J.: CoRR abs/1512.03385 (2015)
Oquab, M., Bottou, L., Laptev, I., Sivic, J.: Learning and transferring mid-level image representations using convolutional neural networks. In: The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2014
Werbos, P.J.: Beyond regression: new tools for prediction and analysis in the behavioral sciences, Ph.D. thesis, Harvard University. Department of Applied Mathematics (1974)
Rumelhart, D.E., Hinton, G.E., Williams, R.J.: Cogn. Model. 5, 1 (1988)
Lowe, D.G.: Object recognition from local scale-invariant features. In: Proceedings of the Seventh IEEE International Conference on Computer Vision (1999)
McConnell, R.: Method of and apparatus for pattern recognition, 28 January 1986. US Patent 4,567,610
Maini, R., Aggarwal, H.: Int. J. Image Process. (IJIP) 3, 1 (2009)
Perfilieva, I., Danková, M., Bede, B.: Towards f-transform of a higher degree. In: IFSA/EUSFLAT Conference (2009)
LeCun, Y., Bottou, L., Bengio, Y., Haffner, P.: Proc. IEEE 86, 2278 (1998)
Fukushima, K.: Biol. Cybern. 36, 193 (1980)
Williams, D.R.G.H.R., Hinton, G.: Nature 323, 533 (1986)
Perfilieva, I., Kreinovich, V.: Fuzzy Sets Syst. 180, 55 (2011)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing AG
About this paper
Cite this paper
Molek, V., Perfilieva, I. (2017). Convolutional Neural Networks with the F-transform Kernels. In: Rojas, I., Joya, G., Catala, A. (eds) Advances in Computational Intelligence. IWANN 2017. Lecture Notes in Computer Science(), vol 10305. Springer, Cham. https://doi.org/10.1007/978-3-319-59153-7_35
Download citation
DOI: https://doi.org/10.1007/978-3-319-59153-7_35
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-59152-0
Online ISBN: 978-3-319-59153-7
eBook Packages: Computer ScienceComputer Science (R0)