Abstract
In recent years, there has been a growing need for techniques that enable users to explain the basis for decisions made by neural networks in image recognition problems. While many conventional methods have focused on presenting the spatial basis for pattern recognition judgments by identifying the regions in the image that significantly influence the recognition results, our proposed method focuses on presenting a structural basis for recognizing handwritten Kanji characters. This is accomplished by pruning neural networks to acquire detectors for simple patterns that are commonly found in Kanji characters. During the process of sequentially adding these detectors, we also apply pruning to the network connecting the detectors, thereby aiming to precisely acquire simple hierarchical connections among the detectors as a structural recognition process in pattern recognition. We successfully applied this method to simple handwritten Kanji images, and achieved Kanji recognition by combining detectors for simple patterns without significantly affecting the recognition rate.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Singh, A., Sengupta, S., Lakshminarayanan, V.: Explainable deep learning models in medical image analysis. J. Imaging 6(6), 52 (2020)
Xu, F., Uszkoreit, H., Du, Y., Fan, W., Zhao, D., Zhu, J.: Explainable AI: A brief survey on history, research areas, approaches and challenges. In: Tang, J., Kan, M.-Y., Zhao, D., Li, S., Zan, H. (eds.) NLPCC 2019. LNCS (LNAI), vol. 11839, pp. 563–574. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-32236-6_51
Sundararajan, M., Taly, A., Yan, Q.: Axiomatic attribution for deep networks. In: International Conference on Machine Learning, pp. 3319–3328 (2017)
Ribeiro, M.T., Singh, S., Guestrin, C.: “Why should I trust you?” Explaining the predictions of any classifier. In: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp. 1135–1144 (2016)
Erhan, D., Bengio, Y., Courville, A., Vincent, P.: Visualizing higher-layer features of a deep network. Univ. Montreal 1341(3), 1 (2009)
Nguyen, A., Yosinski, J., Clune, J.: Understanding neural networks via feature visualization: a survey. In: Samek, W., Montavon, G., Vedaldi, A., Hansen, L.K., Müller, K.-R. (eds.) Explainable AI: Interpreting, Explaining and Visualizing Deep Learning. LNCS (LNAI), vol. 11700, pp. 55–76. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-28954-6_4
Zhang, J., Du, J., Dai, L.: Radical analysis network for learning hierarchies of Chinese characters. Pattern Recogn. 103, 107305 (2020)
Liang, T., Glossner, J., Wang, L., Shi, S., Zhang, X.: Pruning and quantization for deep neural network acceleration: a survey. Neurocomputing 461, 370–403 (2021)
Otsu, N.: A threshold selection method from gray-level histograms. IEEE Trans. Syst. Man Cybernet. 9(1), 62–66 (1979)
Acknowledgements
This work was supported by JSPS KAKENHI (Grant Number JP 19K12045).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Gyohten, K., Ohki, H., Takami, T. (2023). Character Structure Analysis by Adding and Pruning Neural Networks in Handwritten Kanji Recognition. In: Lu, H., Blumenstein, M., Cho, SB., Liu, CL., Yagi, Y., Kamiya, T. (eds) Pattern Recognition. ACPR 2023. Lecture Notes in Computer Science, vol 14407. Springer, Cham. https://doi.org/10.1007/978-3-031-47637-2_10
Download citation
DOI: https://doi.org/10.1007/978-3-031-47637-2_10
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-47636-5
Online ISBN: 978-3-031-47637-2
eBook Packages: Computer ScienceComputer Science (R0)