Abstract
AbstractIn recent years, computer vision and convolutional neural networks have been gradually applied in embedded devices. However, due to the limitation of hardware, the inference speed of many high-precision algorithms is very slow, which requires high performance hardware. In this study, a lightweight network called LightCSPNet is proposed for image classification and object detection. LightCSPNet is built by stacking four identical modules, each of which has adopted an improved CSP (Cross-Stage-Partial-connections) structure for channel number expansion. The special inverse residual structure is constructed for feature extraction, and the transformer modules are added in the proposed model. In this study, the typical defect detection in industry is adopted as testing platform, and a defect dataset consisting of 12 categories including cloth, road, bridge, steel and etc., was constructed for image classification. Compared with MobileNetV3, our model has almost the same accuracy, but the number of parameters and GFLOPs (Giga Floating-point Operations Per Second) have been, respectively, reduced to 88% and 36% for ImageNet100 and the dataset we built. In addition, compared with MobileNetV2 and MobileNetV3 for VOC2012 dataset in object detection, LightCSPNet obtained 0.4% and 0.6% mAP (Mean Average Precision) improvement respectively, and the inference speed on CPU was twice as fast.
Funder
Shandong Provincial Key Laboratory of Precision Manufacturing and Non-traditional Machining
SDUT&Zhangdian District Integration Development Project
Publisher
Springer Science and Business Media LLC
Subject
Computational Mathematics,General Computer Science
Reference35 articles.
1. Hinton, G., Vinyals, O.: Distilling the Knowledge in a Neural Network. In: NIPS Deep Learning and Representation Learning Workshop. (2015).
2. Liu, Z. Li, J., Shen, Z., Huang, G., Yan, S., Zhang, C.: Learning Efficient Convolutional Networks through Network Slimming. In: 2017 IEEE International Conference on Computer Vision (ICCV). pp. 2755–2763 IEEE, Venice (2017). https://doi.org/10.1109/ICCV.2017.298.
3. Gholami, A., Kim, S., Dong, Z., Yao, Z., Mahoney, M.W.: A Survey of Quantization Methods for Efficient Neural Network Inference, http://arxiv.org/abs/2103.13630, (2021).
4. Krizhevsky, A., Sutskever, I., Hinton, G.E.: ImageNet classification with deep convolutional neural networks. Commun. ACM. 60(6), 84–90 (2017). https://doi.org/10.1145/3065386
5. Singh, P., Verma, V.K., Rai, P., Namboodiri, V.P.: HetConv: Heterogeneous Kernel-Based Convolutions for Deep CNNs. In: 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). pp. 4830–4839 (2019). https://doi.org/10.1109/CVPR.2019.00497.
Cited by
2 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献
1. OBGESS: Automating Original Bender Gestalt Test Based on One Stage Deep Learning;International Journal of Computational Intelligence Systems;2023-11-13
2. Comparative Analysis of Lightweight Pre-Trained CNN Models for Coffee Bean Roasting Level Identification;2023 2nd International Conference on Computer System, Information Technology, and Electrical Engineering (COSITE);2023-08-02