Pub Date : 2022-10-01DOI: 10.1109/SmartCloud55982.2022.00016
Heng He, Jiaqi Liu, J. Gu, Feng Gao
Recently multi-cloud has become the main model of cloud computing. With the rapid development of cloud computing technology, users are increasingly concerned about data security in the cloud. To ensure data security, users encrypt private data and upload it to cloud servers. Nevertheless, it is challenging to search ciphertexts with keywords from large amounts of encrypted data of multiple cloud servers. Moreover, existing attribute-based searchable encrypted schemes have several limitations, such as inflexible access control policy, only supporting single or conjunctive keyword search, and low search efficiency. Therefore, we propose an efficient Attribute-based Multi-keyword Search scheme (AMSE) over Encrypted data in multi-cloud environment. AMSE leverages the high-performance Ciphertext-Policy Attribute-Based Encryption (CP-ABE) algorithm to achieve multi-keyword ciphertext search and fine-grained access control. By introducing a retrieval server, AMSE can efficiently and accurately search ciphertexts in multi-cloud. The security analysis and performance evaluation demonstrate that AMSE is secure, highly efficient, and well-suited for multi-cloud.
{"title":"An Efficient Multi-Keyword Search Scheme over Encrypted Data in Multi-Cloud Environment","authors":"Heng He, Jiaqi Liu, J. Gu, Feng Gao","doi":"10.1109/SmartCloud55982.2022.00016","DOIUrl":"https://doi.org/10.1109/SmartCloud55982.2022.00016","url":null,"abstract":"Recently multi-cloud has become the main model of cloud computing. With the rapid development of cloud computing technology, users are increasingly concerned about data security in the cloud. To ensure data security, users encrypt private data and upload it to cloud servers. Nevertheless, it is challenging to search ciphertexts with keywords from large amounts of encrypted data of multiple cloud servers. Moreover, existing attribute-based searchable encrypted schemes have several limitations, such as inflexible access control policy, only supporting single or conjunctive keyword search, and low search efficiency. Therefore, we propose an efficient Attribute-based Multi-keyword Search scheme (AMSE) over Encrypted data in multi-cloud environment. AMSE leverages the high-performance Ciphertext-Policy Attribute-Based Encryption (CP-ABE) algorithm to achieve multi-keyword ciphertext search and fine-grained access control. By introducing a retrieval server, AMSE can efficiently and accurately search ciphertexts in multi-cloud. The security analysis and performance evaluation demonstrate that AMSE is secure, highly efficient, and well-suited for multi-cloud.","PeriodicalId":104366,"journal":{"name":"2022 IEEE 7th International Conference on Smart Cloud (SmartCloud)","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126646425","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-01DOI: 10.1109/SmartCloud55982.2022.00015
Xiuli Li, Guoshi Wang, Chuping Wang, Yan-Fang Qin, Ning Wang
Source code security audit is an effective technique to deal with security vulnerabilities and software bugs. As one kind of white-box testing approaches, it can effectively help developers eliminate defects in the code. However, it suffers from performance issues. In this paper, we propose an incremental checking mechanism which enables fast source code security audits. And we conduct comprehensive experiments to verify the effectiveness of our approach.
{"title":"Software Source Code Security Audit Algorithm Supporting Incremental Checking","authors":"Xiuli Li, Guoshi Wang, Chuping Wang, Yan-Fang Qin, Ning Wang","doi":"10.1109/SmartCloud55982.2022.00015","DOIUrl":"https://doi.org/10.1109/SmartCloud55982.2022.00015","url":null,"abstract":"Source code security audit is an effective technique to deal with security vulnerabilities and software bugs. As one kind of white-box testing approaches, it can effectively help developers eliminate defects in the code. However, it suffers from performance issues. In this paper, we propose an incremental checking mechanism which enables fast source code security audits. And we conduct comprehensive experiments to verify the effectiveness of our approach.","PeriodicalId":104366,"journal":{"name":"2022 IEEE 7th International Conference on Smart Cloud (SmartCloud)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122464168","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
With the rapid development of artificial intelligence, natural language processing, as an important branch, has also become a hot research field. A series of super large-scale pre-trained models represented by BERT and GPT have made great progress in natural language understanding and natural language generation, even some of the experimental accuracy exceed the human benchmark. However, these models will also make some mistakes and even fairness problems when they have the language ability equivalent to human beings. In order to verify whether the models can truly understand natural language, the evaluation of these models is particularly important. More methods are needed to evaluate the model. The language model-based evaluation tools often require a lot of computing resources. In this paper, we propose a method for testing and evaluation of Chinese natural language processing in cloud, generate testing data and design tests for Chinese data and test two pre-trained models. The experimental results show that our method can find defects of the model, though it has high performance on specific dataset.
{"title":"Data Generation, Testing and Evaluation of Chinese Natural Language Processing in the Cloud","authors":"Minjie Ding, Mingang Chen, Wenjie Chen, Lizhi Cai, Yuanhao Chai","doi":"10.1109/SmartCloud55982.2022.00020","DOIUrl":"https://doi.org/10.1109/SmartCloud55982.2022.00020","url":null,"abstract":"With the rapid development of artificial intelligence, natural language processing, as an important branch, has also become a hot research field. A series of super large-scale pre-trained models represented by BERT and GPT have made great progress in natural language understanding and natural language generation, even some of the experimental accuracy exceed the human benchmark. However, these models will also make some mistakes and even fairness problems when they have the language ability equivalent to human beings. In order to verify whether the models can truly understand natural language, the evaluation of these models is particularly important. More methods are needed to evaluate the model. The language model-based evaluation tools often require a lot of computing resources. In this paper, we propose a method for testing and evaluation of Chinese natural language processing in cloud, generate testing data and design tests for Chinese data and test two pre-trained models. The experimental results show that our method can find defects of the model, though it has high performance on specific dataset.","PeriodicalId":104366,"journal":{"name":"2022 IEEE 7th International Conference on Smart Cloud (SmartCloud)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128363880","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-01DOI: 10.1109/SmartCloud55982.2022.00040
Xinyu Wang, Xin Du, Wenli Li, Zhihui Lu
Content Delivery Network (CDN) can store workflow content by deploying edge service nodes and realizing network optimization with reasonable scheduling strategies. Manual selection of appropriate service clusters is coupled with low timeliness and a high economic burden. Therefore, the timing prediction of bandwidth remains a persistent demand to choose service clusters that can take on a safe workload. In this study, we proposes a model which can predict the load level in the future period to optimize the content delivery network. This model uses a machine learning method (K-means) for data optimization, which discarded 3114 of 185572 pieces of data that impact subsequent prediction models. Afterward, the model uses a 4-layers Long Short-Term Memory Network(LSTM) to predict the aggregated temporal data. The model named BK-LSTM considers the execution time and accuracy, eventually learning the real-time bandwidth demand pattern of 654 servers in the specific cluster. Experiments show that our BK-LSTM model has a mean absolute percentage error(MAPE) metric of about 15.2% on the test set, demonstrating this model’s ability to predict bandwidth workload well.
{"title":"A Bandwidth Prediction Method Based on Hybrid LSTM for Content Delivery Network","authors":"Xinyu Wang, Xin Du, Wenli Li, Zhihui Lu","doi":"10.1109/SmartCloud55982.2022.00040","DOIUrl":"https://doi.org/10.1109/SmartCloud55982.2022.00040","url":null,"abstract":"Content Delivery Network (CDN) can store workflow content by deploying edge service nodes and realizing network optimization with reasonable scheduling strategies. Manual selection of appropriate service clusters is coupled with low timeliness and a high economic burden. Therefore, the timing prediction of bandwidth remains a persistent demand to choose service clusters that can take on a safe workload. In this study, we proposes a model which can predict the load level in the future period to optimize the content delivery network. This model uses a machine learning method (K-means) for data optimization, which discarded 3114 of 185572 pieces of data that impact subsequent prediction models. Afterward, the model uses a 4-layers Long Short-Term Memory Network(LSTM) to predict the aggregated temporal data. The model named BK-LSTM considers the execution time and accuracy, eventually learning the real-time bandwidth demand pattern of 654 servers in the specific cluster. Experiments show that our BK-LSTM model has a mean absolute percentage error(MAPE) metric of about 15.2% on the test set, demonstrating this model’s ability to predict bandwidth workload well.","PeriodicalId":104366,"journal":{"name":"2022 IEEE 7th International Conference on Smart Cloud (SmartCloud)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124699736","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-01DOI: 10.1109/SmartCloud55982.2022.00039
Lin Chen, Qingchun Yu, W. Liang, Jiahong Cai, Hangyu Zhu, Songyou Xie
With the increasing growth of electronic medical data, the difficulties of data sharing among medical institutions and the leakage of data privacy have become the focus of the public and medical workers. The blockchain has the characteristics of decentralization, traceability, and immutability, which can provide new ideas for fine-grained secure access to medical research. This article first introduces blockchain and blockchain-based privacy protection technology; then analyzes the advantages and disadvantages of electronic medical records, and introduces the current development status of electronic medical records based on blockchain technology; then from data encryption, access the three aspects of control and transaction anonymity introduce the medical data privacy protection method based on blockchain technology; finally, the full text is summarized and prospected.
{"title":"Overview of Medical Data Privacy Protection based on Blockchain Technology","authors":"Lin Chen, Qingchun Yu, W. Liang, Jiahong Cai, Hangyu Zhu, Songyou Xie","doi":"10.1109/SmartCloud55982.2022.00039","DOIUrl":"https://doi.org/10.1109/SmartCloud55982.2022.00039","url":null,"abstract":"With the increasing growth of electronic medical data, the difficulties of data sharing among medical institutions and the leakage of data privacy have become the focus of the public and medical workers. The blockchain has the characteristics of decentralization, traceability, and immutability, which can provide new ideas for fine-grained secure access to medical research. This article first introduces blockchain and blockchain-based privacy protection technology; then analyzes the advantages and disadvantages of electronic medical records, and introduces the current development status of electronic medical records based on blockchain technology; then from data encryption, access the three aspects of control and transaction anonymity introduce the medical data privacy protection method based on blockchain technology; finally, the full text is summarized and prospected.","PeriodicalId":104366,"journal":{"name":"2022 IEEE 7th International Conference on Smart Cloud (SmartCloud)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127104868","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
For detecting and warning abnormal transaction of virtual cryptocurrency: we proposed PROTECTION (PRivacy-preserving suspiciOus Transaction detECTION), and proposed big matrix inversion algorithm to solve the problem that the physics of TEE is easily limited by memory size. Based on the privacy protection framework, we proposed three supervised learning algorithms to detect and warn abnormal transactions, they respectively are the federated logistic regression model(VERTIGO) over vertically partitioned data, the federated random forest model over vertically partitioned data, and the federated multilayer perceptron model over vertically partitioned data. According to the experimental results, we found that among the three algorithms, the federated logistic regression model(VERTIGO) over vertically partitioned data is ahead of the federated random forest model over vertically partitioned data, and the federated multilayer perceptron model over vertically partitioned data in all indicators, it has a good effect on detecting abnormal transaction of virtual cryptocurrency.
{"title":"Detecting and Warning Abnormal Transaction of Virtual Cryptocurrency Based on Privacy Protection Framework","authors":"Tong Zhu, Chenyang Liao, Lanting Guo, Ziyang Zhou, Wenwen Ruan, Wenhao Wang, Xinyu Li, Qingfu Zhang, Hao Zheng, Shuang Wang, Yuetong Liu","doi":"10.1109/SmartCloud55982.2022.00018","DOIUrl":"https://doi.org/10.1109/SmartCloud55982.2022.00018","url":null,"abstract":"For detecting and warning abnormal transaction of virtual cryptocurrency: we proposed PROTECTION (PRivacy-preserving suspiciOus Transaction detECTION), and proposed big matrix inversion algorithm to solve the problem that the physics of TEE is easily limited by memory size. Based on the privacy protection framework, we proposed three supervised learning algorithms to detect and warn abnormal transactions, they respectively are the federated logistic regression model(VERTIGO) over vertically partitioned data, the federated random forest model over vertically partitioned data, and the federated multilayer perceptron model over vertically partitioned data. According to the experimental results, we found that among the three algorithms, the federated logistic regression model(VERTIGO) over vertically partitioned data is ahead of the federated random forest model over vertically partitioned data, and the federated multilayer perceptron model over vertically partitioned data in all indicators, it has a good effect on detecting abnormal transaction of virtual cryptocurrency.","PeriodicalId":104366,"journal":{"name":"2022 IEEE 7th International Conference on Smart Cloud (SmartCloud)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126777688","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
As an innovative economic form, the digital economy takes data as a new production factor and becomes a new driving force to promote the high-quality development of economy and society. Building a fully digital trading platform plays a very important role in improving the development level and core competitiveness of China’s digital economy.At this stage, there are many problems in data trading, such as insufficient ecosystem, unclear data ownership, imperfect system and mechanism, imperfect supervision system, and shortage of industry professionals.This paper proposes a new generation of Intelligent Data Trading System of niDts, which realizes the liberalization of data trading, providing efficient, convenient, transparent, and safe data product trading services for data circulation transactions. The system has stimulated the multiplier effect of data elements and led to the digital transformation in the fields of economy, life, and governance.
{"title":"niDts: A New Generation Intelligent Data Trading System","authors":"Qifeng Tang, Zhiqing Shao, Lihua Huang, Hsunfang Cho, Yiguang Zhang","doi":"10.1109/SmartCloud55982.2022.00030","DOIUrl":"https://doi.org/10.1109/SmartCloud55982.2022.00030","url":null,"abstract":"As an innovative economic form, the digital economy takes data as a new production factor and becomes a new driving force to promote the high-quality development of economy and society. Building a fully digital trading platform plays a very important role in improving the development level and core competitiveness of China’s digital economy.At this stage, there are many problems in data trading, such as insufficient ecosystem, unclear data ownership, imperfect system and mechanism, imperfect supervision system, and shortage of industry professionals.This paper proposes a new generation of Intelligent Data Trading System of niDts, which realizes the liberalization of data trading, providing efficient, convenient, transparent, and safe data product trading services for data circulation transactions. The system has stimulated the multiplier effect of data elements and led to the digital transformation in the fields of economy, life, and governance.","PeriodicalId":104366,"journal":{"name":"2022 IEEE 7th International Conference on Smart Cloud (SmartCloud)","volume":"63 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126339417","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Transwarp ArgoDB is a one-stop distributed flash database newly issued by Transwarp. It can replace the Hadoop + MPP hybrid architecture, satisfying business’s various demands for big data platform. Therefore, businesses can use big data platform more efficiently and make better use of big data’s commercial value. It supports standard SQL syntax and provides advanced technical capabilities such as multi-mode analysis, realtime data processing, storage and calculation decoupling, mixed load, data federation, mixed deployment of heterogeneous servers and so on. Through an ArgoDB database, we can meet the needs of data warehouse, real-time data warehouse, data mart and federal computing. While reducing platform complexity and it total cost of ownership, it improves business response speed. As an excellent database product, it has successfully replaced Oracle, DB2, Teradata and other foreign products in all walks of life.
{"title":"Transwarp ArgoDB: A Distributed Flash Database","authors":"Changchun Zhang, Cheng Lv, Zhenqiang Chen, Yucheng Lu, Yuxuan Tian, Jiabao Wu, Hongshan Yang","doi":"10.1109/SmartCloud55982.2022.00043","DOIUrl":"https://doi.org/10.1109/SmartCloud55982.2022.00043","url":null,"abstract":"Transwarp ArgoDB is a one-stop distributed flash database newly issued by Transwarp. It can replace the Hadoop + MPP hybrid architecture, satisfying business’s various demands for big data platform. Therefore, businesses can use big data platform more efficiently and make better use of big data’s commercial value. It supports standard SQL syntax and provides advanced technical capabilities such as multi-mode analysis, realtime data processing, storage and calculation decoupling, mixed load, data federation, mixed deployment of heterogeneous servers and so on. Through an ArgoDB database, we can meet the needs of data warehouse, real-time data warehouse, data mart and federal computing. While reducing platform complexity and it total cost of ownership, it improves business response speed. As an excellent database product, it has successfully replaced Oracle, DB2, Teradata and other foreign products in all walks of life.","PeriodicalId":104366,"journal":{"name":"2022 IEEE 7th International Conference on Smart Cloud (SmartCloud)","volume":"66 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124716010","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-01DOI: 10.1109/SmartCloud55982.2022.00036
X. Peng, Yilun Song, Kangjian Yuan, Xu Guo, Zhihui Lu, Jie Liu
Cloud computing is a relatively mature business computing model, which is gradually developed from technologies such as distributed computing, parallel processing, and grid computing. Similarly, with the continuous emergence of cloud computing applications, people’s understanding of cloud computing is also constantly changing. This paper designs a private cloud platform called “UCloudStack” based on cloud computing technology. The platform can provide a complete set of cloud resource management capabilities such as unified management of core services such as virtualization, SDN network, and distributed storage, resource scheduling, monitoring logs, and operation and maintenance, helping the digital transformation of government and enterprises.
{"title":"UCloudStack - A Private Cloud Platform for Lightweight Delivery","authors":"X. Peng, Yilun Song, Kangjian Yuan, Xu Guo, Zhihui Lu, Jie Liu","doi":"10.1109/SmartCloud55982.2022.00036","DOIUrl":"https://doi.org/10.1109/SmartCloud55982.2022.00036","url":null,"abstract":"Cloud computing is a relatively mature business computing model, which is gradually developed from technologies such as distributed computing, parallel processing, and grid computing. Similarly, with the continuous emergence of cloud computing applications, people’s understanding of cloud computing is also constantly changing. This paper designs a private cloud platform called “UCloudStack” based on cloud computing technology. The platform can provide a complete set of cloud resource management capabilities such as unified management of core services such as virtualization, SDN network, and distributed storage, resource scheduling, monitoring logs, and operation and maintenance, helping the digital transformation of government and enterprises.","PeriodicalId":104366,"journal":{"name":"2022 IEEE 7th International Conference on Smart Cloud (SmartCloud)","volume":"121 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122910569","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-10-01DOI: 10.1109/smartcloud55982.2022.00028
ZhanTi Ll, Ni Jia, Hongmei Jin
Aiming at the problem of low image exposure in low-light scenes at night, resulting in low accuracy of fatigue driving detection, a lightweight Zero-DCE night fatigue driving detection algorithm was proposed The depthwise separable convolution is used in the backbone feature extraction nebv0rk of the Zero-DCE model to improve the speed of the detection nebv0rk and reduce the amount of nebv0rk parameters; the down-sampled input is used as the input of the enhanced nebv0rk, and the output is mapped back to the original resolution by up-sampling. Perf0rm image enhancement, effectively balancing enhancement performance and significantly reducing computational cost. The facial eye and mouth features are detected by the target detection algorithm and the open and closed states are identified and the detection results are calculated and output according to the eye and mouth fatigue parameters combined with the threshold The experimental results show that in the low-light environment at night, the detection algorithm proposed in this paper improves the detection accuracy by 17.07% compared with the existing algorithm, and the detection time after algorithm fusion is 0.012s, which is more in line with the application requirements of fatigue driving detection scenarios.
{"title":"Night Fatigue Driving Detection Algorithm based on Lightweight Zero-DCE","authors":"ZhanTi Ll, Ni Jia, Hongmei Jin","doi":"10.1109/smartcloud55982.2022.00028","DOIUrl":"https://doi.org/10.1109/smartcloud55982.2022.00028","url":null,"abstract":"Aiming at the problem of low image exposure in low-light scenes at night, resulting in low accuracy of fatigue driving detection, a lightweight Zero-DCE night fatigue driving detection algorithm was proposed The depthwise separable convolution is used in the backbone feature extraction nebv0rk of the Zero-DCE model to improve the speed of the detection nebv0rk and reduce the amount of nebv0rk parameters; the down-sampled input is used as the input of the enhanced nebv0rk, and the output is mapped back to the original resolution by up-sampling. Perf0rm image enhancement, effectively balancing enhancement performance and significantly reducing computational cost. The facial eye and mouth features are detected by the target detection algorithm and the open and closed states are identified and the detection results are calculated and output according to the eye and mouth fatigue parameters combined with the threshold The experimental results show that in the low-light environment at night, the detection algorithm proposed in this paper improves the detection accuracy by 17.07% compared with the existing algorithm, and the detection time after algorithm fusion is 0.012s, which is more in line with the application requirements of fatigue driving detection scenarios.","PeriodicalId":104366,"journal":{"name":"2022 IEEE 7th International Conference on Smart Cloud (SmartCloud)","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129582806","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：