Abstract
As modern convolutional neural networks become increasingly deeper, they also become slower and require high computational resources beyond the capabilities of many mobile and embedded platforms. To address this challenge, much of the recent research has focused on reducing the model size and computational complexity. In this paper, we propose a novel residual depth-separable convolution block, which is an improvement of the basic building block of MobileNets. We modified the original block by adding an identity shortcut connection (with zero-padding for increasing dimensions) from the input to the output. We demonstrated that the modified architecture with the width multiplier (\(\alpha \)) set to 0.92 slightly outperforms the accuracy and inference time of the baseline MobileNet (\(\alpha = 1\)) on the challenging Places365 dataset while reducing the number of parameters by 14%.
This work has been partially supported by Statutory Funds of Electronics, Telecommunications and Informatics Faculty, Gdansk University of Technology.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Bianco, S., Cadene, R., Celona, L., Napoletano, P.: Benchmark analysis of representative deep neural network architectures. IEEE Access 6 (2018). https://doi.org/10.1109/ACCESS.2018.2877890
Brzeski, A., Grinholc, K., Nowodworski, K., Przybylek, A.: Evaluating performance and accuracy improvements for attention-OCR. In: 18th International Conference on Computer Information Systems and Industrial Management Applications (CISIM 2019), Belgrade, Serbia (2019)
Byra, M., et al.: Impact of ultrasound image reconstruction method on breast lesion classification with neural transfer learning (2018). arXiv:1804.02119
Cychnerski, J., Brzeski, A., Boguszewski, A., Marmolowski, M., Trojanowicz, M.: Clothes detection and classification using convolutional neural networks. In: 22nd IEEE International Conference on Emerging Technologies and Factory Automation (ETFA), Limassol, Cyprus (2017)
Gholami, A., et al.: SqueezeNext: hardware-aware neural network design. In: ECV Workshop at CVPR 2018, Utah, USA (2018)
Han, D., Kim, J., Kim, J.: Deep pyramidal residual networks. In: 30th IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI (2017)
He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: 29th IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV (2016)
He, K., Zhang, X., Ren, S., Sun, J.: Identity mappings in deep residual networks. In: Leibe, B., Matas, J., Sebe, N., Welling, M. (eds.) ECCV 2016. LNCS, vol. 9908, pp. 630–645. Springer, Cham (2016). https://doi.org/10.1007/978-3-319-46493-0_38
Howard, A.G., et al.: MobileNets: efficient convolutional neural networks for mobile vision applications (2017). arXiv:1704.04861
Iandola, F.N., Han, S., Moskewicz, M.W., Ashraf, K., Dally, W.J., Keutzer, K.: SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and \(<\)0.5Â MB model size (2016). arXiv:1602.07360
Jaderberg, M., Vedaldi, A., Zisserman, A.: Speeding up convolutional neural networks with low rank expansions. In: 2014 British Machine Vision Conference, Nottingham, UK (2014)
Janczyk, K., Czuszynski, K., Ruminski, J.: Digits recognition with quadrant photodiode and convolutional neural network. In: 11th International Conference on Human System Interaction (HSI 2018), Gdansk, Poland (2018)
Krizhevsky, A., Sutskever, I., Hinton, G.E.: ImageNet classification with deep convolutional neural networks. In: Advances in Neural Information Processing Systems, pp. 1097–1105 (2012)
Mehta, S., Rastegari, M., Shapiro, L., Hajishirzi, H.: ESPNetv2: a light-weight, power efficient, and general purpose convolutional neural network (2018). arXiv:1811.11431
Podlodowski, L., Roziewski, S., Nurzynski, M.: An ensemble of deep convolutional neural networks for marking hair follicles on microscopic images. In: 2018 Federated Conference on Computer Science and Information Systems (FedCSIS 2018), Poznan, Poland (2018). https://doi.org/10.15439/2018F389
Przybylek, K., Shkroba, I.: Crowd counting á la Bourdieu. In: Workshop on Modern Approaches in Data Engineering and Information System Design at ADBIS 2019, Bled, Slovenia (2019)
Redmon, J., Divvala, S., Girshick, R., Farhadi, A.: You only look once: unified, real-time object detection. In: 29th IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV (2016)
Sandler, S., Howard, A., Zhu, M., Zhmoginov, A., Chen, L.: MobileNetV2: inverted residuals and linear bottlenecks (2018). arXiv:1801.04381
Siam, M., Gamal, M., AbdelRazek, M., Yogomain, S., Jagersand, M., Zhang, H.: A comparative study of real-time semantic segmentation for autonomous driving. In: 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, Utah, USA (2018)
Zhang, X., Zhou, X., Lin, M., Sun, J.: ShuffleNet: an extremely efficient convolutional neural network for mobile devices (2017). arXiv:1707.01083
Zhou, B., Lapedriza, A., Khosla, A., Oliva, A., Torralba, A.: Places: a 10 million image database for scene recognition. IEEE Trans. Pattern Anal. Mach. Intell. 40(6), 1452–1464 (2018)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2019 Springer Nature Switzerland AG
About this paper
Cite this paper
Brzeski, A., Grinholc, K., Nowodworski, K., Przybylek, A. (2019). Residual MobileNets. In: Welzer, T., et al. New Trends in Databases and Information Systems. ADBIS 2019. Communications in Computer and Information Science, vol 1064. Springer, Cham. https://doi.org/10.1007/978-3-030-30278-8_33
Download citation
DOI: https://doi.org/10.1007/978-3-030-30278-8_33
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-30277-1
Online ISBN: 978-3-030-30278-8
eBook Packages: Computer ScienceComputer Science (R0)