Parallel Concatenated Network with Cross-layer Connections for Image Recognition

  • Peng Li
  • Pinqun JiangEmail author
  • Shangyou Zeng
  • Rui Fan
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11434)


The traditional convolutional neural networks are heavy with millions of parameters and the classification accuracy is not high. To address this issue, we propose a novel model called parallel concatenated convolutional neural network with cross-layer connections. The model mainly includes parallel processing and concatenate operation. In parallel processing, the diversity of features is increased by using different sizes of convolution kernels. The parallel outputs are integrated together by concatenate operation. Meanwhile, an improved cross-layer connection structure is also added to the model. At the experimental stage, the model was tested on the Caltech-256 and Food-101 datasets, the experiment results indicate that the constructed PCNet (without cross-layer connections) increases the recognition accuracy by 2.54% and 7.31% compared to AlexNet, and the proposed RPCNet (with cross-layer connections) is improved by 6.12% and 12.28% compared to AlexNet.


Convolution neural network Parallel processing Cross-layer connections Image classification 



Authors acknowledge support of the National Natural Science Foundation of China (Grant Nos. 11465004). Authors are also thankful to the anonymous reviewers whose constructive suggestions helped improve and clarify this manuscript.


  1. 1.
    Deng, L., Yu, D.: Deep learning: methods and applications. Found. Trends Signal Process. 7(3), 197–387 (2014)MathSciNetCrossRefGoogle Scholar
  2. 2.
    Bossard, L., Guillaumin, M., Van Gool, L.: Food-101 – Mining Discriminative Components with Random Forests. In: Fleet, D., Pajdla, T., Schiele, B., Tuytelaars, T. (eds.) ECCV 2014. LNCS, vol. 8694, pp. 446–461. Springer, Cham (2014). Scholar
  3. 3.
    LeCun, Y., Bottou, L., Bengio, Y., Haffner, P.: Gradient-based learning applied to document recognition. Proce. IEEE 86(11), 2278–2324 (1998)CrossRefGoogle Scholar
  4. 4.
    Krizhevsky, A., Sutskever, I., Hinton, G.: Imagenet classification with deep convolutional neural networks. In: NIPS, pp. 1097–1105 (2012)Google Scholar
  5. 5.
    Lin, M., Chen, Q., Yan, S.: Network in network. arXiv preprint (2013) arXiv:1312.4400
  6. 6.
    Szegedy, C., Liu, W., Jia, Y., et al.: Going deeper with convolutions. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 1–9 (2015)Google Scholar
  7. 7.
    He, K., Zhang, X., Ren, S., et al.: Deep residual learning for image recognition. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 770–778 (2016)Google Scholar
  8. 8.
    Simonyan, K., Zisserman, A.: Very deep convolutional networks for large-scale image recognition. arXiv preprint (2014). arXiv:1409.1556
  9. 9.
    Huang, G., Liu, Z., Maaten, L.V.D., et al.: Densely connected convolutional networks. In: 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 2261–2269. IEEE (2017)Google Scholar
  10. 10.
    Howard, A.G., Zhu, M., Chen, B., et al.: Mobilenets: efficient convolutional neural networks for mobile vision applications. arXiv preprint (2014). arXiv:1704.04861
  11. 11.
    Coates, A., Ng, A., Lee, H.: An analysis of single-layer networks in unsupervised feature learning. In: Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics, pp. 215–223 (2011)Google Scholar
  12. 12.
    Ioffe, S., Szegedy, C.: Batch normalization: accelerating deep network training by reducing internal covariate shift. arXiv preprint (2015). arXiv:1502.03167
  13. 13.
    Szegedy, C., Ioffe, S., Vanhoucke, V., et al.: Inception-v4, inception-resnet and the impact of residual connections on learning. In: 31st AAAI Conference on Artificial Intelligence, pp. 4278–4284. AAAI Press, San Francisco (2017)Google Scholar
  14. 14.
    Jia, Y., Shelhamer, E., Donahue, J., et al.: Caffe: convolutional architecture for fast feature embedding. In: Proceedings of the 22nd ACM international conference on Multimedia, pp. 675–678. ACM (2014)Google Scholar
  15. 15.
    Attokaren, D.J., Fernandes, I.G., Sriram, A., et al.: Food classification from images using convolutional neural networks. In: TENCON 2017 IEEE Region 10 Conference, pp. 2801–2806. IEEE (2017)Google Scholar
  16. 16.
    Yanai, K., Kawano, Y.: Food image recognition using deep convolutional network with pre-training and fine-tuning. In: 2015 IEEE International Conference on Multimedia & Expo Workshops (ICMEW), pp. 1–6. IEEE (2015)Google Scholar
  17. 17.
    Pandey, P., Deepthi, A., Mandal, B., et al.: FoodNet: recognizing foods using ensemble of deep networks. IEEE Signal Process. Lett. 24(12), 1758–1762 (2017)CrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.College of Electronic EngineeringGuangxi Normal UniversityGuilinChina

Personalised recommendations