[1]潘俊霖.轻量化卷积神经网络的人脸识别方法[J].大众科技,2022,24(06):18-21.
　Face Recognition Method Based on Lightweight Convolutional Neural Network[J].Popular Science & Technology,2022,24(06):18-21.
点击复制

轻量化卷积神经网络的人脸识别方法()
分享到：

《大众科技》[ISSN:1008-1151/CN:45-1235/N]

卷:
24
期数:
2022年06
页码:
18-21
栏目:
信息技术与通信
出版日期:
2022-06-20

文章信息/Info

Title:
Face Recognition Method Based on Lightweight Convolutional Neural Network
作者:
潘俊霖　
（广西旅岛高速公路服务区有限公司，广西 南宁 530000）
关键词:
人脸识别; 深度学习; 轻量化; 卷积神经网络; 嵌入式设备
Keywords:
face recognition deep learning lightweight convolutional neural network embedded device
文献标志码:
A
摘要:
由于计算资源的局限性，现有基于深度学习技术的人脸识别方法难以部署在嵌入式设备上。针对此问题，文章在现有算法的基础上，提出了一种轻量化卷积神经网络的人脸识别方法。该方法依据分层结构，将像素拼接成高维的人脸表示，通过在Wider Face数据集上进行网络的训练与评测。所提出的识别网络模型大小约为2.5 MB，识别速率约为23 ms，能够满足计算资源有限的嵌入式硬件设备的需求，人脸识别方法同时考虑了嵌入式硬件设备以及卷积神经网络的优缺点，为在各种场景下部署高精度低成本的人脸识别方法提供了一种思路。
Abstract:
Due to the limitation of computing resources, the existing face recognition methods based on deep learning technology are difficult to deploy on embedded devices. To solve this problem, based on the existing algorithms, this paper proposes a lightweight convolutional neural network face recognition method. According to the hierarchical structure, the pixels are spliced into a high-dimensional face representation, and the network is trained and evaluated on the Wider Face dataset. The size of the proposed recognition network model is about 2.5 MB and the recognition rate is about 23 ms, which can meet the needs of embedded hardware devices with limited computing resources. The face recognition method considers the advantages and disadvantages of embedded hardware devices and convolutional neural network, which provides an idea for deploying high-precision and low-cost face recognition methods in various scenes.

参考文献/References:

[1] 王旬，汪淑贤，唐璐丹，等. 一种OpenCV 与云计算结合的人脸识别方法[J]. 大众科技，2019，21(5): 14-15, 48. [2] 卢宏涛，张秦川. 深度卷积神经网络在计算机视觉中的应用研究综述[J]. 数据采集与处理，2016，31(1): 17. [3] Zeiler M, Fergus R. Visualizing and understanding convolutional neural networks[C]. European Conference on Computer Vision, 2013. [4] Taigman Y, Yang M, Ranzato M A, et al. DeepFace: closing the gap to human-level performance in face verification[C]. IEEE Conference on Computer Vision and Pattern Recognition, 2014. [5] Huang G B, MRattar M, Berg T, et al. Labeled faces in the wild: a database for studying face recognition in unconstrained environments[J]. Month, 2008, 3: 321923. [6] Simonyan K, Zisserman A. Very deep convolutional networks for large-scale image recognition[C]. International Conference on Learning Representations, 2015. [7] Szegedy C, Liu W, Jia Y, et al. Going deeper with convolutions[J]. IEEE Computer Society, 2014. [8] He K, Zhang X, Ren S, et al. Deep residual learning for image recognition[C]. Computer Vision and Pattern Recognition, 2016. [9] Tan M, Le Q V. EfficientNet: rethinking model scaling for convolutional neural networks[J]. arXiv e-prints, 2019, 5: 11946. [10] Wikanningrum A, RIachmad R F, Ogata K. Improving lightweight convolutional neural network for facial expression recognition via transfer learning[C]. 2019 International Conference on Computer Engineering, Network, and Intelligent Multimedia (CENIM), 2019. [11] Deng J, Guo J, Zafeiriou S. ArcFace: additive angular margin loss for deep face recognition[J]. arXiv e-prints, 2018, 18: 7698. [12] Wang H, Wang Y, Zhou Z, et al. CosFace: large margin cosine loss for deep face recognition[C]. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2018. [13] Torralb A, Fergus R, Freeman W T. 80 million tiny images: a large data set for nonparametric object and scene recognition[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2008, 30(11): 1958-1970. [14] Jia D, Wei D, Socher R, et al. ImageNet: a large-scale hierarchical image database[C]. 2009 IEEE Conference on Computer Vision and Pattern Recognition, 2009. [15] Yang S, Luo P, Loy C C, et al. WIDER FACE: A Face Detection Benchmark[C]. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016.

相似文献/References:

[1]赵 璞 袁 华.基于稀疏表示的人脸识别算法[J].大众科技,2014,16(04):32.
[2]韩 进 秦宏超 杨颖超 刘文武.基于OpenCV 的嵌入式智能门禁系统设计[J].大众科技,2015,15(03):1.
[3]王 旬,汪淑贤,唐璐丹,等.一种 OpenCV 与云计算结合的人脸识别方法[J].大众科技,2019,21(05):14.
　A Method of Face Recognition based on OpenCV and Cloud Computing[J].Popular Science & Technology,2019,21(06):14.
[4]王宏福　李继文.人脸识别技术在高校安保工作中的应用研究[J].大众科技,2020,22(10):7.
　Research on the Application of Face Recognition Technology in University Security Work[J].Popular Science & Technology,2020,22(06):7.
[5]杨秀霞 陆巍巍 张 毅 闫 实.深度学习视域下研究生学科基础课程应用型教学改革探索[J].大众科技,2021,23(7):100.
　Exploration on the Applied Teaching Reform of the Basic Courses of Postgraduate Subject from the Perspective of Deep Learning[J].Popular Science & Technology,2021,23(06):100.
[6]刘　洁　周明霞　李　敏　王志伟　马金凤　朱　红.基于Unet的多模态脑胶质瘤病灶自动获取模型[J].大众科技,2021,23(11):21.
　Automated Acquisition Model of Multimodal Glioma Lesions Based on Unet[J].Popular Science & Technology,2021,23(06):21.
[7]王曾龙　蒋勇勇 彭海峰 杨　钰 段元梅 郭宜娟　王林惠.一种基于Faster-RCNN的棉花虫害识别与统计方法[J].大众科技,2023,25(5):5.
　A Method for Cotton Pests Identification and Statistics Based on Faster-RCNN[J].Popular Science & Technology,2023,25(06):5.
[8]方丹娜 黄国航 冯振俭 冯博超 高 睿.基于地理信息系统的无人机航飞影像地物识别样本库建设技术研究[J].大众科技,2023,25(11):1.
　Research on the Construction Technology of a Sample Library for UAV Aerial Image Ground-Object Recognition Based on Geographic Information System[J].Popular Science & Technology,2023,25(06):1.
[9]马　深　王　睿　南楚希　吴　静　黄　骥.智慧资产管理平台App端设计[J].大众科技,2023,25(11):24.
　Design of Smart Asset Management Platform at App Terminal[J].Popular Science & Technology,2023,25(06):24.

备注/Memo

备注/Memo:
【收稿日期】2022-02-26 【作者简介】潘俊霖（1991－），男，供职于广西旅岛高速公路服务区有限公司，研究方向为人工智能。
更新日期/Last Update: 2022-07-25