Pub Date : 2022-12-02DOI: 10.1109/ICFTIC57696.2022.10075252
Xuechun Wang, Guifen. Chen
A huge number of computational resource-intensive applications and other content delivery services have evolved with the arrival of the 5G era and the adoption of Internet-connected devices, and the data in the Internet of Vehicles has exhibited accelerated growth. To improve the service performance of telematics, caching content at the network edge is an effective solution to reduce content delivery latency. In this paper, we propose an edge caching scheme based on content popularity with cooperation between base stations, roadside units(RSUs) and vehicles, and use an improved sparrow search algorithm(PSSA)for optimization. Simulations show that the proposed caching scheme has advantages in terms of hit rate improvement and delay reduction.
{"title":"Research on Cache Algorithm for Internet of Vehicles Based on Edge Computing","authors":"Xuechun Wang, Guifen. Chen","doi":"10.1109/ICFTIC57696.2022.10075252","DOIUrl":"https://doi.org/10.1109/ICFTIC57696.2022.10075252","url":null,"abstract":"A huge number of computational resource-intensive applications and other content delivery services have evolved with the arrival of the 5G era and the adoption of Internet-connected devices, and the data in the Internet of Vehicles has exhibited accelerated growth. To improve the service performance of telematics, caching content at the network edge is an effective solution to reduce content delivery latency. In this paper, we propose an edge caching scheme based on content popularity with cooperation between base stations, roadside units(RSUs) and vehicles, and use an improved sparrow search algorithm(PSSA)for optimization. Simulations show that the proposed caching scheme has advantages in terms of hit rate improvement and delay reduction.","PeriodicalId":153754,"journal":{"name":"2022 4th International Conference on Frontiers Technology of Information and Computer (ICFTIC)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115681617","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 : 2022-12-02DOI: 10.1109/ICFTIC57696.2022.10075263
Yufei Zhao
As a developing research direction in the field of computer vision, human posture recognition(HPR) involves many related technologies such as pattern recognition, artificial intelligence, image processing and machine vision. This paper focuses on the application of multi feature fusion(MFF) and block sampling(BS) particle filter algorithm(PFA) in HPR, and briefly analyzes the design of human posture data acquisition terminal, the analysis and extraction of human contour features, and the analysis and extraction of stride features; This paper discusses the MFF and BS PFA, and applies it to HPR. Through the statistical experiments of recognition rate(RR) and sensitivity of virtual reality actions, the effectiveness of MFF and BS PFA applied to HPR is verified.
{"title":"Research on human posture recognition based on multi-feature fusion chunk sampling particle filtering algorithm","authors":"Yufei Zhao","doi":"10.1109/ICFTIC57696.2022.10075263","DOIUrl":"https://doi.org/10.1109/ICFTIC57696.2022.10075263","url":null,"abstract":"As a developing research direction in the field of computer vision, human posture recognition(HPR) involves many related technologies such as pattern recognition, artificial intelligence, image processing and machine vision. This paper focuses on the application of multi feature fusion(MFF) and block sampling(BS) particle filter algorithm(PFA) in HPR, and briefly analyzes the design of human posture data acquisition terminal, the analysis and extraction of human contour features, and the analysis and extraction of stride features; This paper discusses the MFF and BS PFA, and applies it to HPR. Through the statistical experiments of recognition rate(RR) and sensitivity of virtual reality actions, the effectiveness of MFF and BS PFA applied to HPR is verified.","PeriodicalId":153754,"journal":{"name":"2022 4th International Conference on Frontiers Technology of Information and Computer (ICFTIC)","volume":"179 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114644849","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 : 2022-12-02DOI: 10.1109/ICFTIC57696.2022.10075197
Xiangtao Jiang, Hancheng Yu, Yuhao Lv, Xin Zhu
Few-shot object detection(FSOD), which conducted to detect novel object based on massive base class samples and few novel classes samples, has gained extensive research interest from academic and industry. Major existing FSOD approaches basically use Faster RCNN(FRCNN) as basic framework. However, the RPN of the FRCNN architecture generates redundant anchor frames, which leads to slow training and consume massive computing resources. In this paper, we choose Sparse RCNN which has a fixed number of anchor frames and good performances in object detection the basic framework. Traditional FSOD training methods customarily use the method of freezing backbone or deleting the head classification branches, which will lead to overfitting of novel class and perform badly on base classes. To solve this problem, we introduced the mutual distillation layer and IOU Mask module in the head and loss of Sparse RCNN respectively. The mutual distillation layer is a multihead structure, which can distill the base class features when training new class samples, and distill the new class features when training base class samples. Experiments on multiple benchmarks show that our framework is significantly superior to other existing methods, and has faster detection speed.
{"title":"Mutually Distilled Sparse RCNN for Few-Shot Object Detection","authors":"Xiangtao Jiang, Hancheng Yu, Yuhao Lv, Xin Zhu","doi":"10.1109/ICFTIC57696.2022.10075197","DOIUrl":"https://doi.org/10.1109/ICFTIC57696.2022.10075197","url":null,"abstract":"Few-shot object detection(FSOD), which conducted to detect novel object based on massive base class samples and few novel classes samples, has gained extensive research interest from academic and industry. Major existing FSOD approaches basically use Faster RCNN(FRCNN) as basic framework. However, the RPN of the FRCNN architecture generates redundant anchor frames, which leads to slow training and consume massive computing resources. In this paper, we choose Sparse RCNN which has a fixed number of anchor frames and good performances in object detection the basic framework. Traditional FSOD training methods customarily use the method of freezing backbone or deleting the head classification branches, which will lead to overfitting of novel class and perform badly on base classes. To solve this problem, we introduced the mutual distillation layer and IOU Mask module in the head and loss of Sparse RCNN respectively. The mutual distillation layer is a multihead structure, which can distill the base class features when training new class samples, and distill the new class features when training base class samples. Experiments on multiple benchmarks show that our framework is significantly superior to other existing methods, and has faster detection speed.","PeriodicalId":153754,"journal":{"name":"2022 4th International Conference on Frontiers Technology of Information and Computer (ICFTIC)","volume":"1143 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121042265","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 : 2022-12-02DOI: 10.1109/ICFTIC57696.2022.10075278
Long Chen, Ling Luo, Xiaoyan Wang, Jian Tang, Xiao Deng
Power digital equipment defect text classification, which is often used as a guide for defect elimination or fault handling, has attracted the attention of researchers. Although various methods have been proposed, the obtained results are undesirable due to the requirement of a large number of labeled training samples of existing methods. However, the annotating process is typically costly and expert knowledge required in electrical field. In order to solve this problem, this paper uses meta-learning to overcome the problem of small sample size and difficulty of scene migration, which is a way to acquire “learning to learn” and learn new tasks quickly based on the knowledge already acquired. Therefore, we model the task of power digital equipment defect texts classification under a Meta-Learning framework in this paper. Specifically, we utilize a hypergraph-based document modeling to enrich the representation learning for entities, and improve the derived prototypes accordingly by prototype network. In addition, the module of relation network is introduced to train a more meaningful distance function for detect type prediction. Comprehensive experiments are conducted on real-world datasets, and the results demonstrate that the proposed framework outperforms existing methods according to F1-score.
{"title":"A Meta-Learning Framework for Predicting Power Digital Equipment Defect Texts via Hypergraph Modeling","authors":"Long Chen, Ling Luo, Xiaoyan Wang, Jian Tang, Xiao Deng","doi":"10.1109/ICFTIC57696.2022.10075278","DOIUrl":"https://doi.org/10.1109/ICFTIC57696.2022.10075278","url":null,"abstract":"Power digital equipment defect text classification, which is often used as a guide for defect elimination or fault handling, has attracted the attention of researchers. Although various methods have been proposed, the obtained results are undesirable due to the requirement of a large number of labeled training samples of existing methods. However, the annotating process is typically costly and expert knowledge required in electrical field. In order to solve this problem, this paper uses meta-learning to overcome the problem of small sample size and difficulty of scene migration, which is a way to acquire “learning to learn” and learn new tasks quickly based on the knowledge already acquired. Therefore, we model the task of power digital equipment defect texts classification under a Meta-Learning framework in this paper. Specifically, we utilize a hypergraph-based document modeling to enrich the representation learning for entities, and improve the derived prototypes accordingly by prototype network. In addition, the module of relation network is introduced to train a more meaningful distance function for detect type prediction. Comprehensive experiments are conducted on real-world datasets, and the results demonstrate that the proposed framework outperforms existing methods according to F1-score.","PeriodicalId":153754,"journal":{"name":"2022 4th International Conference on Frontiers Technology of Information and Computer (ICFTIC)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125004013","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 : 2022-12-02DOI: 10.1109/ICFTIC57696.2022.10075314
Xiaona Kang, S. Shen
This paper mainly describe the design and implementation of the intelligent four-legged robot dog SPRESENSE-base. It tries to simulate the real dog's behavior. It can walk, run, talk with you, and active track. This article take it as example to study how to design and implement the intelligent four-legged robot dog. It use the self-developed Sobot bionic quadruped robot control system; SpotMicroAI open source model structure was used to construct the hardware model; Sony SPRESENSE development board is used for the main board; Using gyroscope for detecting gait; Using ultrasonic wave for detecting obstacles ahead; using laser radar for detecting the environment around; using voice control module for voice recognition; Using image recognition processing technology to realize the active track. After more than one year of research and development, we have realized the mechanical model of the four-legged robot dog; it can follow the command to walk, trot and other gaits; it can automatically avoid obstacles ahead. The system also has the characteristics of high execution efficiency, portability, compatibility, reliability, practicability and easy expansion because of using C/C++ development language.
{"title":"Design and Implement for Intelligent Four-legged Robot Dog SPRESENSE -Base","authors":"Xiaona Kang, S. Shen","doi":"10.1109/ICFTIC57696.2022.10075314","DOIUrl":"https://doi.org/10.1109/ICFTIC57696.2022.10075314","url":null,"abstract":"This paper mainly describe the design and implementation of the intelligent four-legged robot dog SPRESENSE-base. It tries to simulate the real dog's behavior. It can walk, run, talk with you, and active track. This article take it as example to study how to design and implement the intelligent four-legged robot dog. It use the self-developed Sobot bionic quadruped robot control system; SpotMicroAI open source model structure was used to construct the hardware model; Sony SPRESENSE development board is used for the main board; Using gyroscope for detecting gait; Using ultrasonic wave for detecting obstacles ahead; using laser radar for detecting the environment around; using voice control module for voice recognition; Using image recognition processing technology to realize the active track. After more than one year of research and development, we have realized the mechanical model of the four-legged robot dog; it can follow the command to walk, trot and other gaits; it can automatically avoid obstacles ahead. The system also has the characteristics of high execution efficiency, portability, compatibility, reliability, practicability and easy expansion because of using C/C++ development language.","PeriodicalId":153754,"journal":{"name":"2022 4th International Conference on Frontiers Technology of Information and Computer (ICFTIC)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126834053","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 : 2022-12-02DOI: 10.1109/ICFTIC57696.2022.10075152
X. Chen
Proficiency testing (PT) is one of the most important ways to determine a laboratory's testing and calibration capability. Establishing an efficient and accurate PT evaluation program can provide the right guidance for laboratory supervision. The article demonstrates the differential impact of different data processing methods on the evaluation results, using the different ways of handling data measurement units of test results and data processing software as analysis points. The methodological recommendations for the correct selection of data measurement units are presented, and the differences in the evaluation results of data processing software using different quartile calculation algorithms are given. It provides a scientific basis for EMC laboratory capability evaluation and technical references for laboratory capability evaluation by appropriate supervision and management departments.
{"title":"Analysis of the Variability of EMC Laboratory Proficiency Testing Data Processing Methods on the Evaluation Results","authors":"X. Chen","doi":"10.1109/ICFTIC57696.2022.10075152","DOIUrl":"https://doi.org/10.1109/ICFTIC57696.2022.10075152","url":null,"abstract":"Proficiency testing (PT) is one of the most important ways to determine a laboratory's testing and calibration capability. Establishing an efficient and accurate PT evaluation program can provide the right guidance for laboratory supervision. The article demonstrates the differential impact of different data processing methods on the evaluation results, using the different ways of handling data measurement units of test results and data processing software as analysis points. The methodological recommendations for the correct selection of data measurement units are presented, and the differences in the evaluation results of data processing software using different quartile calculation algorithms are given. It provides a scientific basis for EMC laboratory capability evaluation and technical references for laboratory capability evaluation by appropriate supervision and management departments.","PeriodicalId":153754,"journal":{"name":"2022 4th International Conference on Frontiers Technology of Information and Computer (ICFTIC)","volume":"92 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115018093","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 : 2022-12-02DOI: 10.1109/ICFTIC57696.2022.10075243
Xinyan Zhang
Aiming at the problem that the trust-based access control model cannot dynamically adapt to the changes of the network environment in the Internet of Things (IoT) environment, a dynamic access control model based on data security is proposed. Firstly, the method of association rules is used to analyze the user's historical access behavior, extract the user's frequent access path and create an access behavior database. Then, based on real-time access behavior, the real-time status information of users is obtained, and a spatiotemporal sensitivity credibility evaluation module is constructed by using “spatiotemporal slicing”. Finally, the dynamic access authorization management module and the access control rule module are combined to realize dynamic access control based on user access behavior. Experiments show that the proposed model has higher security than the traditional trust-based access control model.
{"title":"Dynamic access control model based on user access behavior in the Internet of Things environment","authors":"Xinyan Zhang","doi":"10.1109/ICFTIC57696.2022.10075243","DOIUrl":"https://doi.org/10.1109/ICFTIC57696.2022.10075243","url":null,"abstract":"Aiming at the problem that the trust-based access control model cannot dynamically adapt to the changes of the network environment in the Internet of Things (IoT) environment, a dynamic access control model based on data security is proposed. Firstly, the method of association rules is used to analyze the user's historical access behavior, extract the user's frequent access path and create an access behavior database. Then, based on real-time access behavior, the real-time status information of users is obtained, and a spatiotemporal sensitivity credibility evaluation module is constructed by using “spatiotemporal slicing”. Finally, the dynamic access authorization management module and the access control rule module are combined to realize dynamic access control based on user access behavior. Experiments show that the proposed model has higher security than the traditional trust-based access control model.","PeriodicalId":153754,"journal":{"name":"2022 4th International Conference on Frontiers Technology of Information and Computer (ICFTIC)","volume":"45 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115030401","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 : 2022-12-02DOI: 10.1109/ICFTIC57696.2022.10075309
Qingcheng Chen
This paper designs an efficient inspection method of trajectory planning and 3D modeling for crane based on Unmanned Aerial Vehicle (UAV). The working environment of crane is harsh and risky. On the other side, the UAV system has a good convenience to avoid obstacle and fly over dangerous environment. the application of the remote sensing mapping technology of UAV in crane has been a potential future development, which will greatly reduce the workload of manual inspection and improve work efficiency, even can complete the project measurement in some inaccessible location. With this Compounding technology of trajectory planning and 3D modeling, it is much easier to find out the defects of steel structure in crane, such as weld and crack, etc. the 3D modeling process for unstructured working environment will be conducive to realize the intelligent inspection functions of UAV, including autonomous flight, trajectory planning, and obstacle avoidance. Finally a field test results show this approach can promote the safety and work efficiency of inspection in crane.
{"title":"Design of trajectory planning and 3D modeling for crane inspection based on UAV","authors":"Qingcheng Chen","doi":"10.1109/ICFTIC57696.2022.10075309","DOIUrl":"https://doi.org/10.1109/ICFTIC57696.2022.10075309","url":null,"abstract":"This paper designs an efficient inspection method of trajectory planning and 3D modeling for crane based on Unmanned Aerial Vehicle (UAV). The working environment of crane is harsh and risky. On the other side, the UAV system has a good convenience to avoid obstacle and fly over dangerous environment. the application of the remote sensing mapping technology of UAV in crane has been a potential future development, which will greatly reduce the workload of manual inspection and improve work efficiency, even can complete the project measurement in some inaccessible location. With this Compounding technology of trajectory planning and 3D modeling, it is much easier to find out the defects of steel structure in crane, such as weld and crack, etc. the 3D modeling process for unstructured working environment will be conducive to realize the intelligent inspection functions of UAV, including autonomous flight, trajectory planning, and obstacle avoidance. Finally a field test results show this approach can promote the safety and work efficiency of inspection in crane.","PeriodicalId":153754,"journal":{"name":"2022 4th International Conference on Frontiers Technology of Information and Computer (ICFTIC)","volume":"522 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115102105","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}
A malicious PDF file detection method based on ensemble learning is proposed to address the problem that malicious PDF files are highly concealed and difficult to detect. In order to efficiently identify malicious PDF files that are highly concealed, the detection range of malicious PDF files by machine learning models is improved by combining the conventional features of PDF files with structural features. The recognition module adopts Stacking method of ensemble learning, adds weighting operation to improve the combination performance of multiple base learners, and finds the best combination of base learners and meta learner through experiments. After experiments, NB, RF and DT are selected as the optimal Stacking model base learners and Logistic Regression as the meta learner. the optimal Stacking model achieves 98.70% accuracy on the test set, which is better than Adaboost model and deep learning DNN model.
{"title":"A Malicious PDF File Detection Method Based on Improved Ensemble Learning Stacking","authors":"Yidan Tang, Jinjin Dong, Yixuan Guo, Yihan Zhou, Feifan Lu, Bo Zhang","doi":"10.1109/ICFTIC57696.2022.10075332","DOIUrl":"https://doi.org/10.1109/ICFTIC57696.2022.10075332","url":null,"abstract":"A malicious PDF file detection method based on ensemble learning is proposed to address the problem that malicious PDF files are highly concealed and difficult to detect. In order to efficiently identify malicious PDF files that are highly concealed, the detection range of malicious PDF files by machine learning models is improved by combining the conventional features of PDF files with structural features. The recognition module adopts Stacking method of ensemble learning, adds weighting operation to improve the combination performance of multiple base learners, and finds the best combination of base learners and meta learner through experiments. After experiments, NB, RF and DT are selected as the optimal Stacking model base learners and Logistic Regression as the meta learner. the optimal Stacking model achieves 98.70% accuracy on the test set, which is better than Adaboost model and deep learning DNN model.","PeriodicalId":153754,"journal":{"name":"2022 4th International Conference on Frontiers Technology of Information and Computer (ICFTIC)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122884541","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 : 2022-12-02DOI: 10.1109/ICFTIC57696.2022.10075310
Weifa Zheng, Peiyu Cheng, Zitao Cai, Yanjun Xiao
The BiGRU model does not consider the weight of features when extracting features. In order to solve this problem, this paper adds the Attention mechanism to the BiGRU hidden layer, uses the feature vector obtained from the BiGRU as the input of the Attention layer, and uses the attention score as the weight of the feature vector, so that the most important features are retained to the greatest extent. In this paper, BiGRU-Attention model is applied to network attack traffic detection, and CIDDS data set is used for model training and testing. Experiments show that BiGRU-Attention model designed in this paper has high accuracy and F1 value.
{"title":"Research on Network Attack Detection Model Based on BiGRU-Attention","authors":"Weifa Zheng, Peiyu Cheng, Zitao Cai, Yanjun Xiao","doi":"10.1109/ICFTIC57696.2022.10075310","DOIUrl":"https://doi.org/10.1109/ICFTIC57696.2022.10075310","url":null,"abstract":"The BiGRU model does not consider the weight of features when extracting features. In order to solve this problem, this paper adds the Attention mechanism to the BiGRU hidden layer, uses the feature vector obtained from the BiGRU as the input of the Attention layer, and uses the attention score as the weight of the feature vector, so that the most important features are retained to the greatest extent. In this paper, BiGRU-Attention model is applied to network attack traffic detection, and CIDDS data set is used for model training and testing. Experiments show that BiGRU-Attention model designed in this paper has high accuracy and F1 value.","PeriodicalId":153754,"journal":{"name":"2022 4th International Conference on Frontiers Technology of Information and Computer (ICFTIC)","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122160973","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：