An effective face recognition system based on Cloud based IoT with a deep learning model

被引:12
作者
Chauhan, Deepika [1 ]
Kumar, Ashok [2 ]
Bedi, Pradeep [3 ]
Athavale, Vijay Anant [4 ]
Veeraiah, D. [5 ]
Pratap, Boppuru Rudra [6 ]
机构
[1] Shivajirao Kadam Inst Technol & Management, Dept Comp Sci & Engn, Indore, Madhya Pradesh, India
[2] MJP Rohilkhand Univ, Fac Legal Studies, Ctr Cyber Law & Policy Res, PG Dept Law,SRNCT, Bareilly, Uttar Pradesh, India
[3] Graph Era Hill Univ, Dept Comp Sci & Engn, Dehra Dun, Uttarakhand, India
[4] Panipat Inst Engn & Technol, Dept Comp Sci & Engn, Panipat, Haryana, India
[5] Lakireddy Bali Reddy Coll Engn A, Dept Comp Sci & Engn, Mylavarant, Andhra Pradesh, India
[6] CHRIST Deemed Univ, Dept Comp Sci & Engn, Bengaluru, Karnataka, India
关键词
Deep learning; IoT; Cloud; Edge computing; Deep neural network; FUSION; IMAGE;
D O I
10.1016/j.micpro.2020.103726
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
As of late, the Internet of Things (IoT) innovation has been utilized in applications, for example, transportation, medical care, video observation, and so on. The quick appropriation and development of IoT in these segments are producing an enormous measure of information. For instance, IoT gadgets, for example, cameras produce various pictures when utilized in medical clinic reconnaissance sees. Here, face acknowledgement is one of the most significant instruments that can be utilized for clinic affirmations, enthusiastic discovery, and identification of patients, location of fake gadgets. patient, and test clinic models. Programmed and shrewd face acknowledgement frameworks are profoundly precise in an overseen climate; notwithstanding, they are less exact in an unmanaged climate. Additionally, frameworks must keep on running on numerous occasions in different applications, for example, insightful wellbeing. This work presents a tree-based profound framework for programmed face acknowledgement in a cloud climate. The inside and out pattern have been proposed to cost less for the PC without focusing on unwavering quality. In the model, the additional size is isolated into a few sections, and a stick is made for each part. The tree is characterized by its branch area and stature. The branches are spoken to by a leftover capacity, which comprises of a twofold layer, a stack game plan, and a non-direct capacity. The proposed technique is assessed in an assortment of generally accessible information bases. An examination of the method is likewise finished with top to bottom craftsmanship models for the eye to eye connection. The aftereffects of the tests indicated that the example was considered to have accomplished a precision of 98.65%, 99.19%, and 95.84%.
引用
收藏
页数:8
相关论文
共 24 条
[21]   Joint and collaborative representation with local adaptive convolution feature for face recognition with single sample per person [J].
Yang, Meng ;
Wang, Xing ;
Zeng, Guohang ;
Shen, Linlin .
PATTERN RECOGNITION, 2017, 66 :117-128
[22]   IoT big data analytics for smart homes with fog and cloud computing [J].
Yassine, Abdulsalam ;
Singh, Shailendra ;
Hossain, M. Shamim ;
Muhammad, Ghulam .
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE, 2019, 91 :563-573
[23]   Face biometric quality assessment via light CNN [J].
Yu, Jun ;
Sun, Kejia ;
Gao, Fei ;
Zhu, Suguo .
PATTERN RECOGNITION LETTERS, 2018, 107 :25-32
[24]   Edge Intelligence in the Cognitive Internet of Things: Improving Sensitivity and Interactivity [J].
Zhang, Yin ;
Ma, Xiao ;
Zhang, Jing ;
Hossain, M. Shamim ;
Muhammad, Ghulam ;
Amin, Syed Umar .
IEEE NETWORK, 2019, 33 (03) :58-64