Kernel-blending connection approximated by a neural network for image classification

Xinxin Liu; Yunfeng Zhang; Fangxun Bao; Kai Shao; Ziyi Sun; Caiming Zhang

doi:10.1007/s41095-020-0181-9

| Sign up

PDF (706.1 KB)

Cite

EndNote(RIS) BibTeX

Collect

Submit Manuscript

Show Outline

Figures (5)

Fig. 1

Fig. 2

Fig. 3

Fig. 4

Fig. 5

Tables (4)

Table 1

Table 2

Table 3

Table 4

Research Article | Open Access

Kernel-blending connection approximated by a neural network for image classification

Xinxin Liu^¹, Yunfeng Zhang^¹(), Fangxun Bao^², Kai Shao^¹, Ziyi Sun^¹, Caiming Zhang^{¹^,²}

1 Shandong University of Finance and Economics, Jinan 250014, China

2 Shandong University, Jinan 250100, China

Show Author Information

Abstract

This paper proposes a kernel-blending connection approximated by a neural network (KBNN) for image classification. A kernel mapping connection structure, guaranteed by the function approximation theorem, is devised to blend feature extraction and feature classification through neural network learning. First, a feature extractor learns features from the raw images. Next, an automatically constructed kernel mapping connection maps the feature vectors into a feature space. Finally, a linear classifier is used as an output layer of the neural network to provide classification results. Furthermore, a novel loss function involving a cross-entropy loss and a hinge loss is proposed to improve the generalizability of the neural network. Experimental results on three well-known image datasets illustrate that the proposed method has good classification accuracy and generalizability.

Keywords

image classification blending neural network function approximation kernel mapping connection generalizability

References

[1]

Cortes,

; V.

Vapnik,

Support-vector networks. Machine Learning Vol. 20, 273-297, 1995.

Google Scholar

[2]

Bagarinao,

; T.

Kurita,

; M.

Higashikubo,

; H.

Inayoshi,

Adapting SVM image classifiers to changes in imaging conditions using incremental SVM: An application to car detection. In: Computer Vision-ACCV 2009. Lecture Notes in Computer Science, Vol. 5996. H.

Zha,

; R.

Taniguchi,

; S.

Maybank,

Eds. Springer Berlin Heidelberg, 363-372, 2010.

[3]

Y. Q.

Guo,

; X. P.

Jia,

; D.

Paull,

Effective sequential classifier training for SVM-based multitemporal remote sensing image classification. arXiv preprint arXiv:1706.04719, 2017.

[4]

G. E.

Hinton,

; S.

Osindero,

; Y. W.

Teh,

A fast learning algorithm for deep belief nets. Neural Computation Vol. 18, No. 7, 1527-1554, 2006.

Google Scholar

[5]

Bengio,

; A.

Courville,

; P.

Vincent,

Representation learning: A review and new perspectives. IEEE Transactions on Pattern Analysis and Machine Intelligence Vol. 35, No. 8, 1798-1828, 2013.

Google Scholar

[6]

LeCun,

; B. E.

Boser,

; J. S.

Denker,

; D.

Henderson,

; R.

Howard,

; W.

Hubbard,

; L. D.

Jackel,

Back propagation applied to handwritten zip code recognition. Neural Computation Vol. 1, No. 4, 541-551, 1989.

Google Scholar

[7]

Eitel,

; J. T.

Springenberg,

; L.

Spinello,

; M.

Riedmiller,

; W.

Burgard,

Multimodal deep learning for robust RGB-D object recognition. arXiv preprint arXiv:1507.06821, 2015.

[8]

W. W.

Shi,

; Y. H.

Gong,

; X. Y.

Tao,

; D.

Cheng,

; N. N.

Zheng,

Fine-grained image classification using modified DCNNs trained by cascaded softmax and generalized large-margin losses

IEEE Transactions on Neural Networks and Learning Systems Vol. 30, No. 3, 683-694, 2018.

Google Scholar

[9]

X. X.

Niu,

; C. Y.

Suen,

A novel hybrid CNN-SVM classifier for recognizing handwritten digits

Pattern Recognition Vol. 45, No. 4, 1318-1325, 2012.

Google Scholar

[10]

Sun,

; J.

Park,

; K.

Kang,

; J.

Hur

Novel hybrid CNN-SVM model for recognition of functional magnetic resonance images. In: Proceedings of the IEEE International Conference on Systems, Man, and Cybernetics, 1001-1006, 2017.

[11]

D. H.

Hubel,

; T. N.

Wiesel,

Receptive fields and functional architecture of monkey striate cortex. The Journal of Physiology Vol. 195, No. 1, 215-243, 1968.

Google Scholar

[12]

Fukushima,

Neocognitron: A self-organizing neural network model for a mechanism of pattern recognition unaffected by shift in position. Biological Cybernetics Vol. 36, No. 4, 193-202, 1980.

Google Scholar

[13]

M. D.

Zeiler,

; R.

Fergus,

Visualizing and understanding convolutional networks. In: Computer Vision - ECCV 2014. Lecture Notes in Computer Science, Vol. 8689. D.

Fleet,

; T.

Pajdla,

; B.

Schiele,

; T.

Tuytelaars,

Eds. Springer Cham, 818-833, 2014.

[14]

Sermanet,

; D.

Eigen,

; X.

Zhang,

; M.

Mathieu,

; R.

Fergus,

; Y.

LeCun,

OverFeat: Integrated recognition, localization and detection using convolutional networks. arXiv preprint arXiv:1312.6229, 2013.

[15]

F. L.

Zhang,

; X.

Wu,

; R.-L.

Li,

; J.

Wang,

; Z. H.

Zheng,

; S. M.

Hu,

Detecting and removing visual distractors for video aesthetic enhancement. IEEE Transactions on Multimedia Vol. 20, No. 8, 1987-1999, 2018.

Google Scholar

[16]

Y. H.

Wen,

; L.

Gao,

; H. B.

Fu,

; F. L.

Zhang,

; S. H.

Xia,

Graph CNNs with motif and variable temporal block for skeleton-based action recognition. In: Proceedings of the AAAI Conference on Artificial Intelligence Vol. 33, 8989-8996, 2019.

[17]

Ioffe,

; C.

Szegedy,

Batch normalization: Accelerating deep network training by reducing internal covariate shift. arXiv preprint arXiv:1502.03167, 2015.

[18]

Lin,

; Q.

Chen,

; S. C.

Yan,

Network in network. arXiv preprint arXiv:1312.4400, 2013.

[19]

Cybenko,

Approximation by superpositions of a sigmoidal function. Mathematics of Control, Signals and Systems Vol. 2, No. 4, 303-314, 1989.

Google Scholar

[20]

LeCun,

; L.

Bottou,

; Y.

Bengio,

; P.

Haffner,

Gradient-based learning applied to document recognition. Proceedings of the IEEE Vol. 86, No. 11, 2278-2324, 1998.

Google Scholar

[21]

Krizhevsky,

; G.

Hinton,

Learning multiple layers of features from tiny images. Master Thesis. University of Toronto, 2009.

[22]

Tang,

Deep learning using support vector machines. arXiv preprint arXiv:1306.0239, 2015.

[23]

W. T.

Wan,

; Y. Y.

Zhong,

; T. P.

Li,

; J. S.

Chen,

Rethinking feature distribution for loss functions in image classification. arXiv preprint arXiv:1803.02988, 2018.

[24]

Lee,

; R.

Grosse,

; R.

Ranganath,

; A. Y.

Ng,

Convolutional deep belief networks for scalable unsupervised learning of hierarchical representations. In: Proceedings of the 26th Annual International Conference on Machine Learning, 609-616, 2009.

[25]

T. H.

Chan,

; K.

Jia,

; S. H.

Gao,

; J. W.

Lu,

; Z. N.

Zeng,

; Y.

Ma,

PCANet: A simple deep learning baseline for image classification? IEEE Transactions on Image Processing Vol. 24, No. 12, 5017-5032, 2015.

Google Scholar

[26]

Hosseini-Asl,

; J. M.

Zurada,

; O.

Nasraoui,

Deep learning of part-based representation of data using sparse autoencoders with nonnegativity constraints. IEEE Transactions on Neural Networks and Learning Systems Vol. 27, No. 12, 2486-2498, 2016.

Google Scholar

[27]

Bristow,

; A.

Eriksson,

; S.

Lucey,

Fast convolutional sparse coding. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 391-398, 2013.

[28]

C. Y.

Xu,

; C. Y.

Lu,

; X. D.

Liang,

; J. B.

Gao,

; W.

Zheng,

; T. J.

Wang,

; S. C.

Yan,

Multi-loss regularized deep neural network. IEEE Transactions on Circuits and Systems for Video Technology Vol. 26, No. 12, 2273-2283, 2016.

Google Scholar

[29]

I. J.

Goodfellow,

; D.

Warde-Farley,

; M.

Mirza,

Courville,

; Y.

Bengio,

Maxout networks. arXiv preprint arXiv:1302.4389, 2013.

[30]

Wan,

; M.

Zeiler,

; S.

Zhang,

; Y.

LeCun,

; R.

Fergus,

Regularization of neural networks using dropconnect. In: Proceedings of the 30th International Conference on Machine Learning, Vol. 28, 1058-1066, 2013.

[31]

Malinowski,

; M.

Fritz,

Learnable pooling regions for image classification. arXiv preprint arXiv:1301.3516, 2013.

[32]

M. D.

Zeiler,

; R.

Fergus,

Stochastic pooling for regularization of deep convolutional neural networks. arXiv preprint arXiv:1301.3557, 2013.

[33]

He,

; X.

Zhang,

; S.

Ren,

; J.

Sun,

Deep residual learning for image recognition. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 770-778, 2016.

Computational Visual Media

Volume 6 Issue 4,
December 2020

Pages 467-476

DOI: 10.1007/s41095-020-0181-9

Cite this article:

Liu X, Zhang Y, Bao F, et al. Kernel-blending connection approximated by a neural network for image classification. Computational Visual Media, 2020, 6(4): 467-476. https://doi.org/10.1007/s41095-020-0181-9

Layer	Type	Kernel	Stride	Padding	Channels
Data	Input	N/A	N/A	N/A	N/A
CONV 1	Convolution	$5 \times 5$	2	SAME	64
POOL 1	Average pooling	$3 \times 3$	2	VALID	64
CONV 2	Convolution	$3 \times 3$	2	SAME	128
CONV 3	Convolution	$3 \times 3$	2	SAME	256
POOL 2	Max pooling	$2 \times 2$	2	VALID	256
GAP	Average pooling	$1 \times 1$	1	VALID	256
Kernel mapping	Fully connected	$1 \times 1$	N/A	N/A	128
Linear output	Output	$1 \times 1$	N/A	N/A	10

Method	DLSVM	Niu and Suen’s	CNN+softmax	CDBM	PCANet	Deep NCAE	Drplu	KBNN
Error	0.87	0.19	0.68	0.82	0.62	2.09	1.04	0.36

Method	DLSVM	ResNst110+L-GM	ML-DNN	NIN	Maxout Networks	Drop-Connect	KBNN
Error	11.9	4.96	8.12	8.81	9.38	9.32	1.54

Method	Stochastic Pooling	Learned Pooling	Maxout Networks	NIN	ML-DNN	ResNet	KBNN
Error	42.51	43.71	38.57	35.68	34.18	28.62	32.71