Abstract
The convolutional neural networks (CNNs) deployed on devices for visual image processing faces the thorny problems on high system real-time requirements and resource consumption. A high-performance Low-res MobileNet model is constructed to effectively alleviate the high computing resources and storage costs in the real-time image processing. The main works are summarized as: (1) To actively match the input of low-resolution feature map, the MobileNetV2 is further optimized by clipping to simplify the network structure and improve the efficiency of image recognition. (2) To improve the classification accuracy, the Inception structure is used to fill the Dwise layer in depthwise separable convolution to extract more abundant low-resolution features; the activation function during the process of increasing the dimension is replaced to avoid the loss of useful information; Inter-layer connection structure is adopted to strengthen the fusion of feature information between layers. (3) To reduce the network scale, the gradually decreasing expansion factors are used to remove the redundant structure of the model. Subsequently, the Low-res MobileNet is validated and evaluated through data sets of different scales. The experimental results show that this model has smaller scale, less computation and higher classification accuracy compared with other CNN models. The model has 0.36 M parameters and 25.46 M floating point of operations (FLOPs), which is easy to deploy to resource-constrained mobile and embedded devices. The model runs at 35 batches per second, and it achieves an accuracy rate of 89.38%, 71.60%, and 87.08% on CIFAR-10, CIFAR-100, and CINIC-10 datasets, respectively, which is basically suitable for real-time image classification task applied in low-resolution application scenarios.
Similar content being viewed by others
References
Bai L, Lyu Y, Huang X (2021) RoadNet-RT: High Throughput CNN Architecture and SoC Design for Real-Time Road Segmentation. In IEEE Transactions on Circuits and Systems I: Regular Papers (vol. 68, no. 2, pp. 704–714) https://doi.org/10.1109/TCSI.2020.3038139
Cheng G, Zhou PC, Han JW (2018) Duplex metric learning for image set classification. IEEE Trans Image Process 27(1):281–292
Cheng G, Yang CY, Yao XW, Guo L, Han JW (2018) When deep learning meets metric learning: remote sensing image scene classification via learning discriminative CNNs. IEEE Trans Geosci Remote Sens 56(5):2811–2821
Chollet F (2017) Xception: deep learning with depthwise separable convolutions. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 1251–1258). https://doi.org/10.1109/cvpr.2017.195
Darlow L N, Crowley E J, Antoniou A, Storkey A (2018) CINIC-10 is not ImageNet or CIFAR-10. arXiv preprint arXiv:1810.03505
Gu K, Xia ZF, Qiao JF, Lin WS (2020) Deep Dual-Channel neural network for image-based smoke detection. IEEE Transactions on Multimedia 22(2):311–323
Gu K, Liu HY, Xia ZF, Qiao JF, Lin WS, Thalmann D (2021) PM2.5 monitoring: use information abundance measurement and wide and deep learning. IEEE Transactions on Neural Networks and Learning Systems 32(10):4278–4290
He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 770–778). https://doi.org/10.1109/CVPR.2016.90
Howard AG, Zhu M, Chen B, Kalenichenko D (2019) MobileNets: efficient convolutional neural networks for Mobile vision applications. Appl Intell 50(1):107–118
Huang G, Liu S, Laurens van der Maaten (2017) CondenseNet: An Efficient DenseNet using Learned Group Convolutions arXiv preprint arXiv: 1711.09224
Huang G, Liu Z, Laurens V D M, et al (2017) Densely connected convolutional networks. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 4700–4708). https://doi.org/10.1109/cvpr.2017.243
Iandola F N, Han S, Moskewicz M W, Ashraf K, Dally W J, Keutzer K (2016) SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size. arXiv preprint arXiv:1602.07360
Jia Y, Shelhamer E, Donahue J, et al (2014) Caffe: convolutional architecture for fast feature embedding. In ACM Conf Multimedia (pp. 675–678). https://doi.org/10.1145/2647868.2654889
Krizhevsky A, Sutskever I, Hinton G (2012) ImageNet classification with deep convolutional neural networks. Adv Neural Inf Proces Syst 25(2):1106–1114
Liao X, Li KD, Zhu XS, Liu KJR (2020) Robust detection of image operator chain with two-stream convolutional neural network. IEEE Journal of Selected Topics in Signal Processing 14(5):955–968
Liao X, Yu YB, Li B, Li ZP, Qin Z (2020) A new payload partition strategy in color image steganography. IEEE Transactions on Circuits and Systems for Video Technology 30(3):685–696
Lin M, Chen Q, Yan S (2014) Network in network. In Int. Conf. Learning Representations (pp:1–10)
Lobov SA, Mikhaylov AN, Shamshin M, Makarov VA, Kazantsev VB (2020) Spatial properties of STDP in a self-learning spiking neural network enable controlling a Mobile robot. Front Neurosci 14:88–98
Ma M N, Zhang X Y, Zheng H T, Sun J (2018) Shufflenet V2: practical guidelines for efficient CNN architecture design. In European Conf Comput Vision (pp:122–138). https://doi.org/10.1007/978-3-030-01264-9_8
Mehta S, Rastegari M, Caspi A, Shapiro L, Hajishirzi H (2018) ESPNet: efficient spatial pyramid of dilated convolutions for semantic segmentation. In European Conf Comput Vision (pp. 561–580). https://doi.org/10.1007/978-3-030-01249-6_34
Roccetti M, Delnevo G, Casini L, Mirri S (2021) An alternative approach to dimension reduction for pareto distributed data: a case study. Journal of Big Data 8:39–62
Sakib S, Fouda MM, Fadlullah ZM, Nasser N, Alasmary W (2021) A proof-of-concept of ultra-edge smart IoT sensor: a continuous and lightweight arrhythmia monitoring approach. IEEE Access 9:26093–26106. https://doi.org/10.1109/ACCESS.2021.3056509
Sandler M, Howard A, Zhu M, Zhmoginov A, Chen L-C (2018) MobileNetV2: inverted residuals and linear bottlenecks. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 4510–4520). https://doi.org/10.1109/CVPR.2018.00474
Shelhamer E, Long J, Darrell T (2015) Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 3431–3440). https://doi.org/10.1109/cvpr.2015.7298965
Simonyan K, Zisserman A (2014) Very deep convolutional networks for large-scale image recognition. Computer Science 9:55–56
Sun YM, Wong AKC, Kamel MS (2009) Classification of imbalanced data: a review. Int J Pattern Recognit Artif Intell 23(4):687–719
Szegedy C, Liu W, Jia Y, et al, (2015) Going deeper with convolutions. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 1–9). https://doi.org/10.1109/CVPR.2015.7298594
Tan M, Le Q (2019) EfficientNet: rethinking model scaling for convolutional neural networks. In proceedings of the 36th international conference on machine learning 97, 6105-6114
Yang SM, Wang J, Deng B, Liu C, Li HY, Fietkiewicz C, Loparo KA (2019) Real-time neuromorphic system for large-scale conductance-based spiking neural networks. Ieee Transactions on Cybernetics 49(7):2490–2503
Yang SM, Deng B, Wang J, Li HY, Lu ML, Che YQ, Wei XL, Loparo KA (2020) Scalable digital neuromorphic architecture for large-scale biophysically meaningful neural network with multi-compartment neurons. Ieee Transactions on Neural Networks and Learning Systems 31(1):148–162
Yang SM, Gao T, Wang J, Deng B, Lansdell B, Linares-Barranco B (2021) Efficient spike-driven learning with dendritic event-based processing. Front Neurosci 15:601109. https://doi.org/10.3389/fnins.2021.601109
Zhang ZL, Sabuncu MR (2018) Generalized cross entropy loss for training deep neural networks with Noisy labels. Neural Information Processing Systems 31:1–11
Zhang X, Zhou X, Lin M, Sun J (2018) ShuffleNet: an extremely efficient convolutional neural network for Mobile devices. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 6848–6856). https://doi.org/10.1109/CVPR.2018.00716
Zhou N, Liang R, Shi W (2021) A lightweight convolutional neural network for real-time facial expression detection. IEEE Access 9:5573–5584. https://doi.org/10.1109/ACCESS.2020.3046715
Acknowledgments
This research work was supported by National Natural Science Foundation of China (61001049) and Beijing Natural Science Foundation (4172010).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Ethical approval
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed consent
Informed consent was obtained from all individual participants included in the study.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Yuan, H., Cheng, J., Wu, Y. et al. Low-res MobileNet: An efficient lightweight network for low-resolution image classification in resource-constrained scenarios. Multimed Tools Appl 81, 38513–38530 (2022). https://doi.org/10.1007/s11042-022-13157-8
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-022-13157-8