Deep Neural Network for Product Classification System with Korean Character Image

  • Authors

    • Jae-Kyung Sung
    • Sang-Min Park
    • Sang-Yun Sin
    • Yung Bok Kim
    2018-08-29
    https://doi.org/10.14419/ijet.v7i3.33.21008
  • Deep Learning, Product Image, Korean Character image, Shopping Mall, Product Classification

  • Abstract

    This paper proposes a product classification system based on deep learning using Korean character images (Hangul) to search for products in the shopping mall. Generally, an online shopping mall customer searches through a category classification or a product name to purchase a product. When the exact product name or category is not clear, the user has to search its name. However, the product image classification is degraded because the product logos and characters in the package often interfere. To solve such problems, we propose a classification system based on Deep Learning using Korean character images. The learning data of this system uses Korean character images of PHD08, a Hangul (Korean-language) database. The experimental is carried out using product names collected on the web. For the performance experiment, 10 categories of online shopping mall are selected and the classification accuracy is measured and compared with the previous systems.

     

  • References

    1. [1] Sung-Ho Cho, Youn-Joon Lee(2016), A Comparative Study on Design of User Interface for O2O (Online to Offline) Coupon Application Service, Master's Thesis, p.15-80, Hongik University

      [2] http://www.elec4.co.kr/article/articleView.asp?idx=18816

      [3] http://www.itworld.co.kr/news/87229#csidx7dd4d034afb87a3a735ca646a16946d

      [4] Kitae Kim, Wonseok Oh, Geunwon Lim, Eunwoo Cha, Minyoung Shin, Jongwoo Kim(2018), The way to make training data for deep learning model to recognize keywords in product catalog image at E-commerce. Bibliographic info: J Intell Inform Syst 2018March: 24(1): 1~23

      [5] Chung, S. H, Goswami. A, Lee. H, & Hu, J(2012), The impact of images on user clicks in product search, In Proceedings of the Twelfth International Workshop on Multimedia Data Mining, pp. 25-33, ACM, Aug. 2012.

      [6] D. Lowe(1999), Object Recognition from Local Scale Invariant Features, In International Conference on Computer Vision, pp. 1150-1157, 1999.

      [7] D. Lowe(2004), Distinctive Image Features from Scale Invariant Keypoints, International Journal of Computer Vision, vol.2, no.60, pp. 91-110, 2004

      [8] Gi-Ryong Choi, Hye-Wuk Jung and Jee-Hyoung Lee(2012), Contents-based Image Retrieval System Design of Shopping, Proceedings of KIIS Spring Conference, Vol. 22, No. 1, 2012.

      [9] Veit. A, Kovacs. B, Bell. S, McAuley. J, Bala. K, & Belongie. S(2015), Learning visual clothing style with heterogeneous dyadic co-occurrences, In Proceedings of the IEEE International Conference on Computer Vision, pp. 4642-4650, 2015.

      [10] DOI: Yeon-gyu Kim, Eui-young Cha(2016), Streamlined GoogLeNet Algorithm Based on CNN for Korean Character Recognition, J. Korea Inst. Inf. Commun. Eng., Vol. 20, No. 9 : 1657~1665, Sep. 2016. http://dx.doi.org/10.6109/jkiice.2016.20.9.1657

      [11] DOI: Seung-Cheol Baek(2016), Fast and All-Purpose Area-Based Imagery Registration Using ConvNets, Journal of KIISE,, Vol. 43, No. 9, pp. 1034-1042, 2016. http://dx.doi.org/10.5626/JOK. 2016.43.9.1034

      [12] Andreas Veit, Balazs Kovacs, Sean Bell, Julian McAuley, Kavita Bala, Serge Belongie(2015), Deep Learning of Binary Hash Codes for Fast Image Retrieval, The IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, pp. 27-35, 2015.

      [13] A. Krizhevsky, I. Sutskever, G. E. Hinton(2012), Imagenet classification with deep convolutional neural networks, Advances in Neural Information Processing Systems 25, pp. 1097–1105, 2012.

  • Downloads

  • How to Cite

    Sung, J.-K., Park, S.-M., Sin, S.-Y., & Bok Kim, Y. (2018). Deep Neural Network for Product Classification System with Korean Character Image. International Journal of Engineering & Technology, 7(3.33), 179-182. https://doi.org/10.14419/ijet.v7i3.33.21008

    Received date: 2018-10-04

    Accepted date: 2018-10-04

    Published date: 2018-08-29