End-to-end interoperability between different vendors is fundamental for network operators to take advantage of all features of SDN. However, the chimney deployment of operator network cannot efficiently support the end-to-end service distribution and centralized maintenance requirements. Therefore, it is important to realize end-to-end service provisioning and agile opening for operators. This paper first introduces the proposed end-to-end orchestration architecture based on cloud-network collaboration. And then we briefly introduce the corresponding microservice architecture and the northbound interface of OTN network and IP network controller. Meanwhile, this paper demonstrates and validates the provisioning of constrained connectivity end-to-end services in multi-domain and multivendor SDOTN networks based on Extended ACTN Interface Model. The demonstration results show that access OTN equipment can decoupled with Metropolitan area network (MAN) OTN equipment, and it can realizes the low-cost and fast access of private line service on the orchestration system.
{"title":"Experimental Demonstration of End-to-End Service Orchestration Architecture based on Northbound API","authors":"Yanxia Tan, Yong Zhang, Yanlei Zheng, Yacheng Liu, Guangquan Wang, Yuefeng Ji","doi":"10.1109/NaNA53684.2021.00046","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00046","url":null,"abstract":"End-to-end interoperability between different vendors is fundamental for network operators to take advantage of all features of SDN. However, the chimney deployment of operator network cannot efficiently support the end-to-end service distribution and centralized maintenance requirements. Therefore, it is important to realize end-to-end service provisioning and agile opening for operators. This paper first introduces the proposed end-to-end orchestration architecture based on cloud-network collaboration. And then we briefly introduce the corresponding microservice architecture and the northbound interface of OTN network and IP network controller. Meanwhile, this paper demonstrates and validates the provisioning of constrained connectivity end-to-end services in multi-domain and multivendor SDOTN networks based on Extended ACTN Interface Model. The demonstration results show that access OTN equipment can decoupled with Metropolitan area network (MAN) OTN equipment, and it can realizes the low-cost and fast access of private line service on the orchestration system.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125637020","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 : 2021-10-01DOI: 10.1109/NaNA53684.2021.00088
Cuicui Niu, Zhengzhi Pan, Wuchao Shi, Hai Liu
Network construction is beneficial to extract global structural information from vector data. However, node-link information is easy to bring the problem of privacy leakage in network construction. Thus, we provided node-link preserving in network construction via overlapping histogram under local differential privacy. First, we present the general model of network construction using overlapping histogram under local differential privacy. Second, we present algorithm of random perturbation-linking in network construction via overlapping histogram to preserve node-link. Finally, our theoretical and experimental results show that the proposed scheme can achieve the tradeoff between node-link protection and network utility. The proposed scheme provides a basic method against node-link leakage in network construction via overlapping histogram.
{"title":"Network Construction Using Overlapping Histogram under Local Differential Privacy","authors":"Cuicui Niu, Zhengzhi Pan, Wuchao Shi, Hai Liu","doi":"10.1109/NaNA53684.2021.00088","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00088","url":null,"abstract":"Network construction is beneficial to extract global structural information from vector data. However, node-link information is easy to bring the problem of privacy leakage in network construction. Thus, we provided node-link preserving in network construction via overlapping histogram under local differential privacy. First, we present the general model of network construction using overlapping histogram under local differential privacy. Second, we present algorithm of random perturbation-linking in network construction via overlapping histogram to preserve node-link. Finally, our theoretical and experimental results show that the proposed scheme can achieve the tradeoff between node-link protection and network utility. The proposed scheme provides a basic method against node-link leakage in network construction via overlapping histogram.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124887374","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}
System administrators need to monitor various metrics (network traffic, NTP offset, etc.) of their internal services in real-time as a way to determine whether anomalies occur in the system. Traditional Spectral Residual (SR) anomaly detection methods do not take into account the interference of certain human factors (e.g., changes in personal preferences) in certain scenarios, i.e., concept drift. In these scenarios, the accuracy of anomaly detection is bound to be affected. In order to guarantee the availability and stability of network services, we propose an intelligent and pervasive anomaly detection strategy, CD-SR. First, we use the traditional SR model and the SVM method to train the time series that have not drifted to determine the threshold value. Then, to solve the problem of the pervasiveness of application scenarios, we use a drift detection model to find the time series where concept drift occurs. Finally, the sequence where the concept drift occurs is imported into the drift adaptation model to complete the replacement of the old and new concepts, the data is processed in real-time, and the replaced data is detected again in the detection model for anomalies. In the experimental stage, we obtained several data metrics using the cloud platform system built by Openstack, and by comparing several mainstream anomaly detection algorithms, our method obtained superior results.
{"title":"CD-SR: A Real-time Anomaly Detection Framework for Continuous Concept Drift","authors":"Zhongyi Ding, Shujie Yang, Zhaoyang Liu, Tengchao Ma, Zichen Feng, Mingze Wang","doi":"10.1109/NaNA53684.2021.00040","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00040","url":null,"abstract":"System administrators need to monitor various metrics (network traffic, NTP offset, etc.) of their internal services in real-time as a way to determine whether anomalies occur in the system. Traditional Spectral Residual (SR) anomaly detection methods do not take into account the interference of certain human factors (e.g., changes in personal preferences) in certain scenarios, i.e., concept drift. In these scenarios, the accuracy of anomaly detection is bound to be affected. In order to guarantee the availability and stability of network services, we propose an intelligent and pervasive anomaly detection strategy, CD-SR. First, we use the traditional SR model and the SVM method to train the time series that have not drifted to determine the threshold value. Then, to solve the problem of the pervasiveness of application scenarios, we use a drift detection model to find the time series where concept drift occurs. Finally, the sequence where the concept drift occurs is imported into the drift adaptation model to complete the replacement of the old and new concepts, the data is processed in real-time, and the replaced data is detected again in the detection model for anomalies. In the experimental stage, we obtained several data metrics using the cloud platform system built by Openstack, and by comparing several mainstream anomaly detection algorithms, our method obtained superior results.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125079718","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 : 2021-10-01DOI: 10.1109/NaNA53684.2021.00045
Mingbo Pan, Weibin Su, Yikai Wang
In the background of edge computing becoming an important technology of China’s new infrastructure construction, it is necessary for colleges and universities to set up a new professional curriculum system for edge computing. It is construction and learning from the refinement of artificial intelligence which is easier to industrialize. Through the interdisciplinary integration and optimization construction of the curriculum, this course system is suitable for universities. According to developed course content, design the teaching schedule, and then, based on this scheme, we will cultivate edge computing research and application-oriented talents, and support the transformation and upgrading of industrial automation to intelligent.
{"title":"Review of Research on the Curriculum for Artificial Intelligence and Industrial Automation based on Edge Computing","authors":"Mingbo Pan, Weibin Su, Yikai Wang","doi":"10.1109/NaNA53684.2021.00045","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00045","url":null,"abstract":"In the background of edge computing becoming an important technology of China’s new infrastructure construction, it is necessary for colleges and universities to set up a new professional curriculum system for edge computing. It is construction and learning from the refinement of artificial intelligence which is easier to industrialize. Through the interdisciplinary integration and optimization construction of the curriculum, this course system is suitable for universities. According to developed course content, design the teaching schedule, and then, based on this scheme, we will cultivate edge computing research and application-oriented talents, and support the transformation and upgrading of industrial automation to intelligent.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127081832","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 : 2021-10-01DOI: 10.1109/NaNA53684.2021.00016
Shiyi Zou, Ling-ge Jiang, Pingping Ji, Chen He, Di He, Guorong Zhang
In this letter, a beam selection algorithm is put forward for beamspace high altitude platform massive multiple-input multiple-output (HAP-MIMO) systems. Specifically, the algorithm is subject to the maximization of match degree between the beams and users, which is constructed by exploiting statistical channel state information (CSI). The beam selection is partitioned into two parts. In the first part, we obtain a reduced-dimensional dominant beam set consisting of each user’s most preferred beams according to the match degree. In the second part, we formulate the selection of optimal beams from the dominant beam set as an assignment problem that can be solved by Kuhn-Munkres algorithm. Numerical results demonstrate the performance enhancement of the proposed algorithm with respect to energy efficiency.
{"title":"Beam Selection Algorithm for Beamspace HAP-MIMO Systems Based on Statistical CSI","authors":"Shiyi Zou, Ling-ge Jiang, Pingping Ji, Chen He, Di He, Guorong Zhang","doi":"10.1109/NaNA53684.2021.00016","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00016","url":null,"abstract":"In this letter, a beam selection algorithm is put forward for beamspace high altitude platform massive multiple-input multiple-output (HAP-MIMO) systems. Specifically, the algorithm is subject to the maximization of match degree between the beams and users, which is constructed by exploiting statistical channel state information (CSI). The beam selection is partitioned into two parts. In the first part, we obtain a reduced-dimensional dominant beam set consisting of each user’s most preferred beams according to the match degree. In the second part, we formulate the selection of optimal beams from the dominant beam set as an assignment problem that can be solved by Kuhn-Munkres algorithm. Numerical results demonstrate the performance enhancement of the proposed algorithm with respect to energy efficiency.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"1443 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127443524","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}
In statistical queries work, such as frequency estimation, the untrusted data collector could as an honest-but-curious (HbC) or malicious adversary to learn true values. Local differential privacy(LDP) protocols have been applied against the untrusted third party in data collecting. Nevertheless, excessive noise of LDP will reduce data utility, thus affecting the results of statistical queries. Therefore, it is significant to research the trade-off between privacy and utility. In this paper, we first measure the privacy loss by observing the maximum posterior confidence of the adversary (data collector). Then, through theoretical analysis and comparison we obtain the most suitable utility measure that is Wasserstein distance. Based on these, we introduce an originality framework for privacy-utility tradeoff framework, finding that this system conforms to the Pareto optimality state and formalizing a payoff function to find optimal equilibrium point under Pareto efficiency. Finally, we illustrate the efficacy of our system model by the Adult dataset from the UCI machine learning repository.
{"title":"The Trade-off Between Privacy and Utility in Local Differential Privacy","authors":"Mengqian Li, Youliang Tian, Junpeng Zhang, Dandan Fan, Dongmei Zhao","doi":"10.1109/NaNA53684.2021.00071","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00071","url":null,"abstract":"In statistical queries work, such as frequency estimation, the untrusted data collector could as an honest-but-curious (HbC) or malicious adversary to learn true values. Local differential privacy(LDP) protocols have been applied against the untrusted third party in data collecting. Nevertheless, excessive noise of LDP will reduce data utility, thus affecting the results of statistical queries. Therefore, it is significant to research the trade-off between privacy and utility. In this paper, we first measure the privacy loss by observing the maximum posterior confidence of the adversary (data collector). Then, through theoretical analysis and comparison we obtain the most suitable utility measure that is Wasserstein distance. Based on these, we introduce an originality framework for privacy-utility tradeoff framework, finding that this system conforms to the Pareto optimality state and formalizing a payoff function to find optimal equilibrium point under Pareto efficiency. Finally, we illustrate the efficacy of our system model by the Adult dataset from the UCI machine learning repository.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"69 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121349315","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 : 2021-10-01DOI: 10.1109/NaNA53684.2021.00009
Wendi Sun, Xiaoying Liu, Kechen Zheng, Yang Xu, Jia Liu
Motivated by the dilemma in the multi-channel spectrum sensing by the multi-antenna user, this paper focuses on the scheme of spectrum sensing and energy harvesting to coordinate the time scheduling and energy management in the cognitive radio networks (CRNs), where a secondary transmitter (ST) exploits spectrum holes to transmit data through multichannel. To improve spectrum utilization efficiency, we investigate how the ST selects channels for sensing based on the residual energy, and adjusts the time scheduling of energy harvesting and data transmission with respect to the sensing results. To address this problem, we propose an adaptive scheme concerning spectrum sensing, channel selection, energy harvesting, and data transmission (SCED) for the ST. Moreover, we formulate the optimization of spectrum utilization efficiency as a Markov decision process (MDP) problem, which is challenging due to the system space and action space. Furthermore, we solve the MDP problem by a proposed value iteration algorithm. Numerical results show that the spectrum utilization efficiency under the SCED scheme is better than that under other schemes.
{"title":"Spectrum Utilization Improvement for Multi-Channel Cognitive Radio Networks with Energy Harvesting","authors":"Wendi Sun, Xiaoying Liu, Kechen Zheng, Yang Xu, Jia Liu","doi":"10.1109/NaNA53684.2021.00009","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00009","url":null,"abstract":"Motivated by the dilemma in the multi-channel spectrum sensing by the multi-antenna user, this paper focuses on the scheme of spectrum sensing and energy harvesting to coordinate the time scheduling and energy management in the cognitive radio networks (CRNs), where a secondary transmitter (ST) exploits spectrum holes to transmit data through multichannel. To improve spectrum utilization efficiency, we investigate how the ST selects channels for sensing based on the residual energy, and adjusts the time scheduling of energy harvesting and data transmission with respect to the sensing results. To address this problem, we propose an adaptive scheme concerning spectrum sensing, channel selection, energy harvesting, and data transmission (SCED) for the ST. Moreover, we formulate the optimization of spectrum utilization efficiency as a Markov decision process (MDP) problem, which is challenging due to the system space and action space. Furthermore, we solve the MDP problem by a proposed value iteration algorithm. Numerical results show that the spectrum utilization efficiency under the SCED scheme is better than that under other schemes.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126667398","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 : 2021-10-01DOI: 10.1109/NaNA53684.2021.00034
Yichuan Wang, He Wang, Xinhong Hei, Wenjiang Ji, Lei Zhu
In recent years, a variety of network attacks emerge in an endless stream, and network attacks gradually show the characteristics of higher secrecy and greater harm. At present, the analysis of system vulnerabilities is generally focused on the characteristics analysis and impact hazard level, and lack of formal modeling and vulnerability analysis methods. In this paper, we model the OpenSSL service’s Heartbleed vulnerability based on Petri net. We build a formal model combined with the source code and system state, analyze the vulnerability of the system running state, and propose an automatic vulnerability repair scheme. Experiments show that this model can carry out fine-grained formal analysis and simulation of the Heartbleed, which is of great significance to explore the system vulnerability modeling method.
{"title":"Petri net modeling and vulnerability analysis of the Heartbleed","authors":"Yichuan Wang, He Wang, Xinhong Hei, Wenjiang Ji, Lei Zhu","doi":"10.1109/NaNA53684.2021.00034","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00034","url":null,"abstract":"In recent years, a variety of network attacks emerge in an endless stream, and network attacks gradually show the characteristics of higher secrecy and greater harm. At present, the analysis of system vulnerabilities is generally focused on the characteristics analysis and impact hazard level, and lack of formal modeling and vulnerability analysis methods. In this paper, we model the OpenSSL service’s Heartbleed vulnerability based on Petri net. We build a formal model combined with the source code and system state, analyze the vulnerability of the system running state, and propose an automatic vulnerability repair scheme. Experiments show that this model can carry out fine-grained formal analysis and simulation of the Heartbleed, which is of great significance to explore the system vulnerability modeling method.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115996748","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 : 2021-10-01DOI: 10.1109/NaNA53684.2021.00057
Guorong Zhang, Ling-ge Jiang, Pingping Ji, Shiyi Zou, Chen He, Di He
In this paper, we propose a new user grouping scheme for the high altitude platform (HAP) massive Multiple-Input Multiple-Output (MIMO) systems based on statistical-eigenmode (SE). It has been proved that SE makes a major contribution to signal power for HAPs. Then, a Fubini-Study distance based modified K-means (FS-MKM) user grouping method is proposed aiming at reducing intra-group interference and improving system performance. The proposed modified K-means algorithm improves the initial points selection of the original K-means algorithm. The Fubini-Study distance is obtained based on the SEs of different users. Simulation results confirm that the proposed user grouping algorithm yields significant performance enhancement.
{"title":"A Modified K-means User Grouping Design for HAP Massive MIMO Systems","authors":"Guorong Zhang, Ling-ge Jiang, Pingping Ji, Shiyi Zou, Chen He, Di He","doi":"10.1109/NaNA53684.2021.00057","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00057","url":null,"abstract":"In this paper, we propose a new user grouping scheme for the high altitude platform (HAP) massive Multiple-Input Multiple-Output (MIMO) systems based on statistical-eigenmode (SE). It has been proved that SE makes a major contribution to signal power for HAPs. Then, a Fubini-Study distance based modified K-means (FS-MKM) user grouping method is proposed aiming at reducing intra-group interference and improving system performance. The proposed modified K-means algorithm improves the initial points selection of the original K-means algorithm. The Fubini-Study distance is obtained based on the SEs of different users. Simulation results confirm that the proposed user grouping algorithm yields significant performance enhancement.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116689568","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 : 2021-10-01DOI: 10.1109/NaNA53684.2021.00056
Sugang Ma, Zixian Zhang, Lei Zhang, Yanping Chen, Xiaobao Yang, Lei Pu, Z. Hou
In an effort to the problem of insufficient tracking performance of the Fully-convolutional Siamese network (SiamFC) in complex scenarios, a dual attention mechanism object tracking algorithm based on the Fully-convolutional Siamese network is proposed to improve the generalization capability of the tracker by ameliorating the robustness of the template characteristics. Firstly, a global context attention module is appended after the backbone network of SiamFC to ameliorate the power of original feature extraction from two dimensions of spatial and channel. Then, a coordinate attention module is introduced to augment the capability of feature extraction in the channel dimension. Finally, the model of the proposed algorithm is trained on the Got-10k dataset. Five related algorithms are tested on the OTB2015 dataset, the results of experiments manifest that our algorithm outperforms the baseline trackers, the success and precision rate of the proposed algorithm are improved by 3.3% and 6.3%. The average tracking speed is 145FPS, which can demand the requirement of real-time tracking.
{"title":"Dual attention mechanism object tracking algorithm based on Fully-convolutional Siamese network","authors":"Sugang Ma, Zixian Zhang, Lei Zhang, Yanping Chen, Xiaobao Yang, Lei Pu, Z. Hou","doi":"10.1109/NaNA53684.2021.00056","DOIUrl":"https://doi.org/10.1109/NaNA53684.2021.00056","url":null,"abstract":"In an effort to the problem of insufficient tracking performance of the Fully-convolutional Siamese network (SiamFC) in complex scenarios, a dual attention mechanism object tracking algorithm based on the Fully-convolutional Siamese network is proposed to improve the generalization capability of the tracker by ameliorating the robustness of the template characteristics. Firstly, a global context attention module is appended after the backbone network of SiamFC to ameliorate the power of original feature extraction from two dimensions of spatial and channel. Then, a coordinate attention module is introduced to augment the capability of feature extraction in the channel dimension. Finally, the model of the proposed algorithm is trained on the Got-10k dataset. Five related algorithms are tested on the OTB2015 dataset, the results of experiments manifest that our algorithm outperforms the baseline trackers, the success and precision rate of the proposed algorithm are improved by 3.3% and 6.3%. The average tracking speed is 145FPS, which can demand the requirement of real-time tracking.","PeriodicalId":414672,"journal":{"name":"2021 International Conference on Networking and Network Applications (NaNA)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115009769","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：