Pub Date : 2021-04-23DOI: 10.1109/ICCCS52626.2021.9449186
Yuting Huang, Fan Bai, Chongwen Wang
The penetration rate of mobile phones and tablet computers among college students is increasing, and the loose teaching environment has led to a large number of phubbers in college classrooms. The state of students' attendance in class is an intuitive indicator of classroom quality. Obtaining this data in real-time will bring great help to school evaluation and improvement of teaching standards. The data in this article comes from teaching videos collected by high-definition cameras in colleges. Through offline training, the face detector HDN can accurately extract the position coordinates of the student in the picture in the real teaching scene and pass the detected head information to the convolutional network responsible for judging the state of the student's head to obtain the student's current Class status. The HDN designed in this paper achieves a recall rate of more than 95% on the authoritative public dataset FDDB, and the accuracy of Wider Face's face dataset under three difficulty conditions is 93.9%, 93.2%, and 88.0%. The self-designed Raised Head Network achieves 88% accuracy on the RaisedHead dataset.
手机和平板电脑在大学生中的普及率越来越高，宽松的教学环境导致了大量的低头族出现在大学教室里。学生出勤情况是课堂教学质量的直观指标。实时获取这些数据对学校评价和教学水平的提高有很大的帮助。本文的数据来源于高校高清摄像机采集的教学视频。人脸检测器HDN通过离线训练，能够在真实教学场景中准确提取出学生在图片中的位置坐标，并将检测到的头部信息传递给负责判断学生头部状态的卷积网络，从而获得学生当前的Class状态。本文设计的HDN在权威公共数据集FDDB上实现了95%以上的查全率，在三种难度条件下对Wider Face人脸数据集的查全率分别为93.9%、93.2%和88.0%。自行设计的raise Head Network在RaisedHead数据集上达到88%的准确率。
{"title":"Research on Head Detection and State Estimation Algorithm in Classroom Scene","authors":"Yuting Huang, Fan Bai, Chongwen Wang","doi":"10.1109/ICCCS52626.2021.9449186","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449186","url":null,"abstract":"The penetration rate of mobile phones and tablet computers among college students is increasing, and the loose teaching environment has led to a large number of phubbers in college classrooms. The state of students' attendance in class is an intuitive indicator of classroom quality. Obtaining this data in real-time will bring great help to school evaluation and improvement of teaching standards. The data in this article comes from teaching videos collected by high-definition cameras in colleges. Through offline training, the face detector HDN can accurately extract the position coordinates of the student in the picture in the real teaching scene and pass the detected head information to the convolutional network responsible for judging the state of the student's head to obtain the student's current Class status. The HDN designed in this paper achieves a recall rate of more than 95% on the authoritative public dataset FDDB, and the accuracy of Wider Face's face dataset under three difficulty conditions is 93.9%, 93.2%, and 88.0%. The self-designed Raised Head Network achieves 88% accuracy on the RaisedHead dataset.","PeriodicalId":376290,"journal":{"name":"2021 IEEE 6th International Conference on Computer and Communication Systems (ICCCS)","volume":"45 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127339523","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 : 2021-04-23DOI: 10.1109/ICCCS52626.2021.9449293
Zhenghao Zhang, Liguo Sun, Lingyu Si, C. Zheng
Unsupervised Video Object Segmentation (UVOS) aims to generate accurate pixel-level masks for moving objects without any prior knowledge. A lot of UVOS methods process frames independently by using image segmentation model without considering the temporal information between consecutive frames. Other works rely on RNNs or motion cues to find objects that need to be tracked, these models learn short-term temporal dependencies and thus tend to accumulate errors over time. We propose a new Siamese Co-attention Network to tackle Unsupervised Video Object Segmentation task based on SOLOv2. The Co-attention module in our Siamese Network captures global correspondences between a reference frame and the current one from same video, and it can learn pairwise correlation at any distance to help current frame correctly distinguish primary objects from a global view. Our proposed method is evaluated in TianChi VOS Challenge and DAVIS2017, and the results indicate that it exhibits superior performance.
{"title":"A New Siamese Co-attention Network for Unsupervised Video Object Segmentation","authors":"Zhenghao Zhang, Liguo Sun, Lingyu Si, C. Zheng","doi":"10.1109/ICCCS52626.2021.9449293","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449293","url":null,"abstract":"Unsupervised Video Object Segmentation (UVOS) aims to generate accurate pixel-level masks for moving objects without any prior knowledge. A lot of UVOS methods process frames independently by using image segmentation model without considering the temporal information between consecutive frames. Other works rely on RNNs or motion cues to find objects that need to be tracked, these models learn short-term temporal dependencies and thus tend to accumulate errors over time. We propose a new Siamese Co-attention Network to tackle Unsupervised Video Object Segmentation task based on SOLOv2. The Co-attention module in our Siamese Network captures global correspondences between a reference frame and the current one from same video, and it can learn pairwise correlation at any distance to help current frame correctly distinguish primary objects from a global view. Our proposed method is evaluated in TianChi VOS Challenge and DAVIS2017, and the results indicate that it exhibits superior performance.","PeriodicalId":376290,"journal":{"name":"2021 IEEE 6th International Conference on Computer and Communication Systems (ICCCS)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117068335","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 : 2021-04-23DOI: 10.1109/ICCCS52626.2021.9449264
Siyou Su, D. Hao, Jian-Bin Huang, T. Zhang, Z. Cao, Ning Zhang
All Fiber Displacement Interferometer System for Any Reflector (AFDISAR) provides outstanding measurement accuracy and resolution for recording the dynamic damage evolution and fracture of materials in the shock wave and detonation physics research. Based on sine curve fitting, a new numerical phase demodulation method is presented for the low SNR AFDISAR Doppler signal while measuring low speed moving target. The phase difference between reference and measurement beam is determined by the displacement of the measurement target and changes the output current of the photodetector. Sine curve fitting method is introduced to modulate the phase from the differential frequency of Doppler signal. Traditional Hilbert transform and new sine curve fitting method are simulated separately for modulating the displacement from the low SNR laser Doppler signals with DC-bias offset. The simulation results show that the decoding accuracy of sine curve fitting method is better than the accuracy of Hilbert transform. When SNR downing to −5dB, the error of sine curve fitting is only 1°, while the value for Hilbert transform is as large as58.19°. The measurement results shows that sine curve fitting method has better robustness and tolerance to noise.
{"title":"Phase Demodulation Method for Low SNR Laser Doppler Signals","authors":"Siyou Su, D. Hao, Jian-Bin Huang, T. Zhang, Z. Cao, Ning Zhang","doi":"10.1109/ICCCS52626.2021.9449264","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449264","url":null,"abstract":"All Fiber Displacement Interferometer System for Any Reflector (AFDISAR) provides outstanding measurement accuracy and resolution for recording the dynamic damage evolution and fracture of materials in the shock wave and detonation physics research. Based on sine curve fitting, a new numerical phase demodulation method is presented for the low SNR AFDISAR Doppler signal while measuring low speed moving target. The phase difference between reference and measurement beam is determined by the displacement of the measurement target and changes the output current of the photodetector. Sine curve fitting method is introduced to modulate the phase from the differential frequency of Doppler signal. Traditional Hilbert transform and new sine curve fitting method are simulated separately for modulating the displacement from the low SNR laser Doppler signals with DC-bias offset. The simulation results show that the decoding accuracy of sine curve fitting method is better than the accuracy of Hilbert transform. When SNR downing to −5dB, the error of sine curve fitting is only 1°, while the value for Hilbert transform is as large as58.19°. The measurement results shows that sine curve fitting method has better robustness and tolerance to noise.","PeriodicalId":376290,"journal":{"name":"2021 IEEE 6th International Conference on Computer and Communication Systems (ICCCS)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121342436","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 : 2021-04-23DOI: 10.1109/ICCCS52626.2021.9449312
Xingyu He, Tao Liu
This paper considers recovery of two-dimensional (2D) sparse signals from incomplete measurements. The 2D sparse signals can be reconstructed by solving a sparse representation problem for Multiple Measurement Vectors (MMV). However, the extension of the sparse recovery algorithms to the MMV case may be inefficient if the vectors do not have the same sparsity profile. In this paper, a sequential sparse recovery (SSR) algorithm is proposed to reconstruct the two-dimensional (2D) sparse matrix. The sparsity of the matrix is much reduced after down-sampling observation and the sparse matrix can be reconstructed after sequential observations and reconstructions. Simulation results verify the effectiveness of the proposed method in 2D sparse signal reconstruction.
{"title":"Sparse Matrix Reconstruction Based on Sequential Sparse Recovery for Multiple Measurement Vectors","authors":"Xingyu He, Tao Liu","doi":"10.1109/ICCCS52626.2021.9449312","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449312","url":null,"abstract":"This paper considers recovery of two-dimensional (2D) sparse signals from incomplete measurements. The 2D sparse signals can be reconstructed by solving a sparse representation problem for Multiple Measurement Vectors (MMV). However, the extension of the sparse recovery algorithms to the MMV case may be inefficient if the vectors do not have the same sparsity profile. In this paper, a sequential sparse recovery (SSR) algorithm is proposed to reconstruct the two-dimensional (2D) sparse matrix. The sparsity of the matrix is much reduced after down-sampling observation and the sparse matrix can be reconstructed after sequential observations and reconstructions. Simulation results verify the effectiveness of the proposed method in 2D sparse signal reconstruction.","PeriodicalId":376290,"journal":{"name":"2021 IEEE 6th International Conference on Computer and Communication Systems (ICCCS)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115584653","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 : 2021-04-23DOI: 10.1109/ICCCS52626.2021.9449176
Xu Zhao, Lvwen Huang, Yanming Nie
It is greatly significant to predict air temperature accurately for effective warning of extreme weather events, whereas the complex and nonlinear characteristics of meteorological data make this kind of forecast difficult to achieve high accuracy. To deal with this issue, a novel model named CNN-GRU-RPASM (Convolution Neural Networks - Gated Recurrent Unit - Relative Position-based Self-Attention Mechanism) was proposed in this paper. Apart from the traditional counterparts, the CNN-GRU-RPASM model innovatively combines the advantages of CNN and GRU, and introduces a gaussian amplifier model to improve the self-attention mechanism with relative position information. Firstly, CNN was used to extract the characteristics of the meteorological input data. Then, the improved self-attention mechanism was employed to extract key information from the data sequence. And finally, GRU was utilized to encode the relationship information among time-series data. The performance evaluation with the real meteorological data shows that the CNN-GRU-RP ASM model performs better than its traditional counterparts. This new model will be deployed in the agricultural production service system to provide technical supports for extreme weather disaster warning forecasting.
{"title":"Temperature Prediction Based on Integrated Deep Learning and Attention Mechanism","authors":"Xu Zhao, Lvwen Huang, Yanming Nie","doi":"10.1109/ICCCS52626.2021.9449176","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449176","url":null,"abstract":"It is greatly significant to predict air temperature accurately for effective warning of extreme weather events, whereas the complex and nonlinear characteristics of meteorological data make this kind of forecast difficult to achieve high accuracy. To deal with this issue, a novel model named CNN-GRU-RPASM (Convolution Neural Networks - Gated Recurrent Unit - Relative Position-based Self-Attention Mechanism) was proposed in this paper. Apart from the traditional counterparts, the CNN-GRU-RPASM model innovatively combines the advantages of CNN and GRU, and introduces a gaussian amplifier model to improve the self-attention mechanism with relative position information. Firstly, CNN was used to extract the characteristics of the meteorological input data. Then, the improved self-attention mechanism was employed to extract key information from the data sequence. And finally, GRU was utilized to encode the relationship information among time-series data. The performance evaluation with the real meteorological data shows that the CNN-GRU-RP ASM model performs better than its traditional counterparts. This new model will be deployed in the agricultural production service system to provide technical supports for extreme weather disaster warning forecasting.","PeriodicalId":376290,"journal":{"name":"2021 IEEE 6th International Conference on Computer and Communication Systems (ICCCS)","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122826761","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}
The design of safety-critical systems must concern both cost and availability. However, the design space of redundancy is large with increasing system scale and complexity. Achieving optimal configurations that balance availability and cost can be difficult in the large design space. Therefore, we propose an optimization method for system architectures using the multi-objective evolutionary algorithm based on constrained decomposition with grids (MOEA-CDG). Firstly, a bi-objective model is defined and the availability is calculated on the basis of the discrete-time Bayesian network (DTBN). Then, MOEA-CDG is used to achieve the optimal configurations that meet both cost and availability. Finally, the proposed method is illustrated with an example of the Integrated Modular Avionics (IMA) core processing system, and the results indicate that the method can improve the efficiency of architecture design and outperforms elitist non-dominated sorting genetic algorithm (NSGA-II).
{"title":"Multi-objective Architecture Optimization Based on Evolutionary Algorithm with Grid Decomposition","authors":"Rui Zhang, Lisong Wang, Xinye Cai, Guonan Cui, Yang Hong, Qin Zhang","doi":"10.1109/ICCCS52626.2021.9449098","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449098","url":null,"abstract":"The design of safety-critical systems must concern both cost and availability. However, the design space of redundancy is large with increasing system scale and complexity. Achieving optimal configurations that balance availability and cost can be difficult in the large design space. Therefore, we propose an optimization method for system architectures using the multi-objective evolutionary algorithm based on constrained decomposition with grids (MOEA-CDG). Firstly, a bi-objective model is defined and the availability is calculated on the basis of the discrete-time Bayesian network (DTBN). Then, MOEA-CDG is used to achieve the optimal configurations that meet both cost and availability. Finally, the proposed method is illustrated with an example of the Integrated Modular Avionics (IMA) core processing system, and the results indicate that the method can improve the efficiency of architecture design and outperforms elitist non-dominated sorting genetic algorithm (NSGA-II).","PeriodicalId":376290,"journal":{"name":"2021 IEEE 6th International Conference on Computer and Communication Systems (ICCCS)","volume":"625 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123065719","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 : 2021-04-23DOI: 10.1109/ICCCS52626.2021.9449187
Jing Jiang, Peizhe Xin, Jun Li, Liu Han, Bin Hou, Lixin He
A two-tier 5g resource allocation model for shared network structure, multi-service, and multi-scenario is proposed. The upper layer is an auction-based network resource allocation algorithm, to maximize the social benefits of auction participants. It will convert users' service requests into corresponding bidding information according to business types and models the slice resource allocation problem as an online winner determination problem based on multi-business. The lower level is to distribute the resources obtained from the upper level to the users of the three scenarios of the smart grid, taking the users' requirements as the primary consideration. While allocating network resources, the iteration method is adopted to avoid resource waste caused by resource allocation exceeding demand. The simulation results show that the proposed architecture has a better performance in social benefits and user satisfaction.
{"title":"Research on Resource Allocation Algorithm of 5G Network in Multi-Business Smart Grid","authors":"Jing Jiang, Peizhe Xin, Jun Li, Liu Han, Bin Hou, Lixin He","doi":"10.1109/ICCCS52626.2021.9449187","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449187","url":null,"abstract":"A two-tier 5g resource allocation model for shared network structure, multi-service, and multi-scenario is proposed. The upper layer is an auction-based network resource allocation algorithm, to maximize the social benefits of auction participants. It will convert users' service requests into corresponding bidding information according to business types and models the slice resource allocation problem as an online winner determination problem based on multi-business. The lower level is to distribute the resources obtained from the upper level to the users of the three scenarios of the smart grid, taking the users' requirements as the primary consideration. While allocating network resources, the iteration method is adopted to avoid resource waste caused by resource allocation exceeding demand. The simulation results show that the proposed architecture has a better performance in social benefits and user satisfaction.","PeriodicalId":376290,"journal":{"name":"2021 IEEE 6th International Conference on Computer and Communication Systems (ICCCS)","volume":"85 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115825000","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 : 2021-04-23DOI: 10.1109/ICCCS52626.2021.9449130
Ruirui Mao
With the continuous development of technology in the artificial intelligence era, smart city applications based on high-performance servers in large data centers have penetrated all walks of life. However, the current mainstream smart city application model is only data collection on the device side, and then calculations and inferences in the data center. Data transmission is difficult to achieve the real-time performance of the system, resulting in poor effects in many smart city applications. The intelligent perception for smart city is required to perceive the whole urban environment comprehensively. Among them, small-size pixel target detection and recognition is particularly critical. To this end, a real-time small-size pixel target perception algorithm based on embedded system for smart city is proposed, which uses lightweight neural networks and model pruning optimization to realize terminal intelligence for smart city applications, and integrates traditional machine learning filtering algorithms for improving the detection speed and the accuracy. The experimental results of the method show that the real-time performance and the accuracy of detection are greatly improved for different sizes of small-size pixel targets.
{"title":"Real-time Small-size Pixel Target Perception Algorithm Based on Embedded System for Smart City","authors":"Ruirui Mao","doi":"10.1109/ICCCS52626.2021.9449130","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449130","url":null,"abstract":"With the continuous development of technology in the artificial intelligence era, smart city applications based on high-performance servers in large data centers have penetrated all walks of life. However, the current mainstream smart city application model is only data collection on the device side, and then calculations and inferences in the data center. Data transmission is difficult to achieve the real-time performance of the system, resulting in poor effects in many smart city applications. The intelligent perception for smart city is required to perceive the whole urban environment comprehensively. Among them, small-size pixel target detection and recognition is particularly critical. To this end, a real-time small-size pixel target perception algorithm based on embedded system for smart city is proposed, which uses lightweight neural networks and model pruning optimization to realize terminal intelligence for smart city applications, and integrates traditional machine learning filtering algorithms for improving the detection speed and the accuracy. The experimental results of the method show that the real-time performance and the accuracy of detection are greatly improved for different sizes of small-size pixel targets.","PeriodicalId":376290,"journal":{"name":"2021 IEEE 6th International Conference on Computer and Communication Systems (ICCCS)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115333992","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 : 2021-04-23DOI: 10.1109/ICCCS52626.2021.9449263
Yunkai Yu, Hong Zhu, Meiyi Xie
With the popularity of smart devices and social applications, vast amounts of trajectory data are generated that can be used for traffic planning, etc. However, when trajectory data are applied in these applications, the private information contained in the trajectories can be revealed. In this paper, we focus on trajectory correlation, which can reveal the social relations of users and further cause severe breaches of privacy. We present a method for correlated trajectory publication with differential privacy, called CTP. First, we discretize the continuous geographical space of raw trajectories to obtain a grid space via an adaptive grid partition method with the Laplace mechanism and convert the trajectories from locations into cells. Then, we quantify the trajectory correlation using the cell visit probability vectors of raw trajectories of the cell mode and turn to reducing the similarity of two cell visit probability vectors for the protection of trajectory correlation. Second, based on the correlations extracted from raw trajectories of the cell mode, we design a constrained optimization problem. By solving it via particle swarm optimization, which is modified to satisfy differential privacy, we can obtain an updated cell visit probability vector of a given trajectory, thus weakening the correlations between the given trajectory and other trajectories. Finally, based on the updated probability vector, we synthesize a trajectory corresponding to the given trajectory. We perform experiments on real trajectory datasets. The experimental results show that CTP is stable and achieves a better trade-off between the data utility and the privacy than the existing methods.
{"title":"CTP: Correlated Trajectory Publication with Differential Privacy","authors":"Yunkai Yu, Hong Zhu, Meiyi Xie","doi":"10.1109/ICCCS52626.2021.9449263","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449263","url":null,"abstract":"With the popularity of smart devices and social applications, vast amounts of trajectory data are generated that can be used for traffic planning, etc. However, when trajectory data are applied in these applications, the private information contained in the trajectories can be revealed. In this paper, we focus on trajectory correlation, which can reveal the social relations of users and further cause severe breaches of privacy. We present a method for correlated trajectory publication with differential privacy, called CTP. First, we discretize the continuous geographical space of raw trajectories to obtain a grid space via an adaptive grid partition method with the Laplace mechanism and convert the trajectories from locations into cells. Then, we quantify the trajectory correlation using the cell visit probability vectors of raw trajectories of the cell mode and turn to reducing the similarity of two cell visit probability vectors for the protection of trajectory correlation. Second, based on the correlations extracted from raw trajectories of the cell mode, we design a constrained optimization problem. By solving it via particle swarm optimization, which is modified to satisfy differential privacy, we can obtain an updated cell visit probability vector of a given trajectory, thus weakening the correlations between the given trajectory and other trajectories. Finally, based on the updated probability vector, we synthesize a trajectory corresponding to the given trajectory. We perform experiments on real trajectory datasets. The experimental results show that CTP is stable and achieves a better trade-off between the data utility and the privacy than the existing methods.","PeriodicalId":376290,"journal":{"name":"2021 IEEE 6th International Conference on Computer and Communication Systems (ICCCS)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115935399","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 : 2021-04-23DOI: 10.1109/ICCCS52626.2021.9449146
Peicong Hu, Wendong Yang, Na Pu, Yunfei Peng, Xiang Ding
In this paper, we simulate digital signals of six usual modulation patterns considering Rician fading and propose a new deep neural network structure (CGDNN) combining Convolutional Neural Networks (CNNs) with Gated Recurrent Unit (GRU). Simulation results show that the proposed structure has the ability to classify the signal modulation patterns regardless the influence of different Rician K-factors and has better performance than conventional structures including CNNs and CLDNNs.
{"title":"A New Deep Architecture for Digital Signal Modulation Classification over Rician Fading","authors":"Peicong Hu, Wendong Yang, Na Pu, Yunfei Peng, Xiang Ding","doi":"10.1109/ICCCS52626.2021.9449146","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449146","url":null,"abstract":"In this paper, we simulate digital signals of six usual modulation patterns considering Rician fading and propose a new deep neural network structure (CGDNN) combining Convolutional Neural Networks (CNNs) with Gated Recurrent Unit (GRU). Simulation results show that the proposed structure has the ability to classify the signal modulation patterns regardless the influence of different Rician K-factors and has better performance than conventional structures including CNNs and CLDNNs.","PeriodicalId":376290,"journal":{"name":"2021 IEEE 6th International Conference on Computer and Communication Systems (ICCCS)","volume":"50 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132409152","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：