Pub Date : 2021-04-23DOI: 10.1109/ICCCS52626.2021.9449155
S. Luo, Heng Li, Shuyuan Jin
The Domain Name System (DNS) is a fundamental Internet service that translates domain names into IP addresses [1]. Information leakage of DNS servers may cause the disclosure of users' behaviour or network zone structure and may even induce phishing or intranet penetration attacks. This paper solves two problems: the first is how to measure the information leakage of the DNS, and the second is how to measure the extent that current DNS services suffer from in-formation leakage. The measuring approach proposed in this paper utilizes DNS lookups on a total of 6,936,431 actual running DNS servers with open DNS ports (accounting for all the DNS servers in IPv4 we can reach), and there was a response rate of 71.84%. The experiment demonstrates that 84.41% of the PTR (Pointer, one DNS record type) RRs (resource records) leak network zone structure or business information. A defense method is further proposed to prevent information leakage to enable DNS privacy improvements.
{"title":"Measuring Information Leakage of DNS Server","authors":"S. Luo, Heng Li, Shuyuan Jin","doi":"10.1109/ICCCS52626.2021.9449155","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449155","url":null,"abstract":"The Domain Name System (DNS) is a fundamental Internet service that translates domain names into IP addresses [1]. Information leakage of DNS servers may cause the disclosure of users' behaviour or network zone structure and may even induce phishing or intranet penetration attacks. This paper solves two problems: the first is how to measure the information leakage of the DNS, and the second is how to measure the extent that current DNS services suffer from in-formation leakage. The measuring approach proposed in this paper utilizes DNS lookups on a total of 6,936,431 actual running DNS servers with open DNS ports (accounting for all the DNS servers in IPv4 we can reach), and there was a response rate of 71.84%. The experiment demonstrates that 84.41% of the PTR (Pointer, one DNS record type) RRs (resource records) leak network zone structure or business information. A defense method is further proposed to prevent information leakage to enable DNS privacy improvements.","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":"121349786","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/ICCCS52626.2021.9449103
Hongxia Sun, Naijie Gu, Chuanwen Lin
Impurity defect plays an important role in tire defects, which may directly affect driving safety. A novel model is proposed by this paper for impurity defect detection in tire X-ray images. Firstly, a binarization algorithm based on column grayscale correction is designed, which can obtain more details than other algorithms and lay a solid foundation for the followup detection. Next, a tire X-ray image segmentation algorithm can segment the tire image accurately. Finally, two thresholds are used to judge whether there is an impurity defect in the tire image. The model is evaluated on a real data set which contains various types of impurity defects and achieve good results. But it should be noted that this model can only detect impurity defect in tire body.
{"title":"Tire Impurity Defect Detection Based on Grayscale Correction and Threading Method","authors":"Hongxia Sun, Naijie Gu, Chuanwen Lin","doi":"10.1109/ICCCS52626.2021.9449103","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449103","url":null,"abstract":"Impurity defect plays an important role in tire defects, which may directly affect driving safety. A novel model is proposed by this paper for impurity defect detection in tire X-ray images. Firstly, a binarization algorithm based on column grayscale correction is designed, which can obtain more details than other algorithms and lay a solid foundation for the followup detection. Next, a tire X-ray image segmentation algorithm can segment the tire image accurately. Finally, two thresholds are used to judge whether there is an impurity defect in the tire image. The model is evaluated on a real data set which contains various types of impurity defects and achieve good results. But it should be noted that this model can only detect impurity defect in tire body.","PeriodicalId":376290,"journal":{"name":"2021 IEEE 6th International Conference on Computer and Communication Systems (ICCCS)","volume":"20 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":"122909791","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/ICCCS52626.2021.9449225
Jinjian Huang, Z. Qiu, Weitao Pan, Jun Li, Zhi-Qiang Gao, Bingbing Han, Zihao Xiong, M. Dong
A high-performance HIMAC (HINOC Media Access Control layer) acceleration coprocessor for High-Performance Network over Coax (HINOC3.0) is presented in this paper. The design aims to provide wire-speed processing of up to 10G Ethernet frames. The design also completes the following functions: 1) Supports 2 to 16 channel bonding of the physical layer HIPHY (HINOC Physical layer); 2) Supports up to 128 terminal devices; 3) Supports the filtering and classification of Ethernet frames, completing operations such as assigning priority, discarding, and redirecting; 4) Supports mutual conversion between EMAC frame and HIMAC frame, by completing the encapsulating and splitting work; 5) Supports data exchange between EMAC interface and HIMAC interface. With the implementation of FPGAs, this HIMAC design can support wire-speed processing of up to 10G Ethernet frames and provide media access control layer functions for HINOC3.0.
{"title":"Design of HIMAC Coprocessor for HINOC3.0","authors":"Jinjian Huang, Z. Qiu, Weitao Pan, Jun Li, Zhi-Qiang Gao, Bingbing Han, Zihao Xiong, M. Dong","doi":"10.1109/ICCCS52626.2021.9449225","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449225","url":null,"abstract":"A high-performance HIMAC (HINOC Media Access Control layer) acceleration coprocessor for High-Performance Network over Coax (HINOC3.0) is presented in this paper. The design aims to provide wire-speed processing of up to 10G Ethernet frames. The design also completes the following functions: 1) Supports 2 to 16 channel bonding of the physical layer HIPHY (HINOC Physical layer); 2) Supports up to 128 terminal devices; 3) Supports the filtering and classification of Ethernet frames, completing operations such as assigning priority, discarding, and redirecting; 4) Supports mutual conversion between EMAC frame and HIMAC frame, by completing the encapsulating and splitting work; 5) Supports data exchange between EMAC interface and HIMAC interface. With the implementation of FPGAs, this HIMAC design can support wire-speed processing of up to 10G Ethernet frames and provide media access control layer functions for HINOC3.0.","PeriodicalId":376290,"journal":{"name":"2021 IEEE 6th International Conference on Computer and Communication Systems (ICCCS)","volume":"76 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":"126608648","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/ICCCS52626.2021.9449304
Yang Cheng, Bin-bing Chen, Shouyang Zhong
Traditional communication station recognition usually needs to manually extract features and then use pattern recognition for classification. There are problems of complex preprocessing and difficult feature extraction. Power amplifiers are the core components of shortwave radio stations, and their fingerprint features can be used as radio stations. For the characteristics of individual recognition, a method of power amplifier fingerprint recognition based on one-dimensional convolutional neural network (1D-CNN) is proposed. The characteristic is that it can directly learn features from the original signal and finally complete the classification and recognition. By establishing a power amplifier model, the shortwave signal can obtain fingerprint characteristics. A four-layer 1D-CNN network is used to extract signal features, the network parameters are optimized through experiments, and finally the Softmax classifier is used for classification and recognition. A 98% recognition rate was obtained through experiments, which verified the effectiveness of the method.
{"title":"Fingerprint Feature Recognition of Power Amplifier Based on One-dimensional Convolutional Neural Network","authors":"Yang Cheng, Bin-bing Chen, Shouyang Zhong","doi":"10.1109/ICCCS52626.2021.9449304","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449304","url":null,"abstract":"Traditional communication station recognition usually needs to manually extract features and then use pattern recognition for classification. There are problems of complex preprocessing and difficult feature extraction. Power amplifiers are the core components of shortwave radio stations, and their fingerprint features can be used as radio stations. For the characteristics of individual recognition, a method of power amplifier fingerprint recognition based on one-dimensional convolutional neural network (1D-CNN) is proposed. The characteristic is that it can directly learn features from the original signal and finally complete the classification and recognition. By establishing a power amplifier model, the shortwave signal can obtain fingerprint characteristics. A four-layer 1D-CNN network is used to extract signal features, the network parameters are optimized through experiments, and finally the Softmax classifier is used for classification and recognition. A 98% recognition rate was obtained through experiments, which verified the effectiveness of the method.","PeriodicalId":376290,"journal":{"name":"2021 IEEE 6th International Conference on Computer and Communication Systems (ICCCS)","volume":"48 4 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":"126119354","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/ICCCS52626.2021.9449142
Zhijiao Guo, Biao Hou, Junxing Zhang
With the popularity of wearable devices, smart wearable devices containing various sensors have been widely adopted in healthcare applications. However, there is little research on the use of these devices to study lifestyle regularity, especially to study lifestyle regularity of college students using physiological or exercise data collected by smart wearable devices. In this work, we use the wrist wearable devices worn by students every day to collect college students' daily routine data, and establish models to analyze the regularity of the collected data and propose the use of MOE (Mixture of Experts) and transfer learning to improve the classification performance of the model. The experimental results show that the classification accuracy can be improved by 8.3% using MOE compared with not using it, and the accuracy can be further increased by 2.9% with Transfer Learning.
{"title":"A Study of College Students' Lifestyle Regularity Based on Wearable Devices and Deep Learning","authors":"Zhijiao Guo, Biao Hou, Junxing Zhang","doi":"10.1109/ICCCS52626.2021.9449142","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449142","url":null,"abstract":"With the popularity of wearable devices, smart wearable devices containing various sensors have been widely adopted in healthcare applications. However, there is little research on the use of these devices to study lifestyle regularity, especially to study lifestyle regularity of college students using physiological or exercise data collected by smart wearable devices. In this work, we use the wrist wearable devices worn by students every day to collect college students' daily routine data, and establish models to analyze the regularity of the collected data and propose the use of MOE (Mixture of Experts) and transfer learning to improve the classification performance of the model. The experimental results show that the classification accuracy can be improved by 8.3% using MOE compared with not using it, and the accuracy can be further increased by 2.9% with Transfer Learning.","PeriodicalId":376290,"journal":{"name":"2021 IEEE 6th International Conference on Computer and Communication Systems (ICCCS)","volume":"90 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":"126066640","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/ICCCS52626.2021.9449292
Xiao Xiao, Xinyu Liu, Dairong Peng, Tong Chen
In recent years, tissue oxygen saturation (StO2) based on hyperspectral imaging (HSI) technology has been deeply studied in the field of affective computing. HSI StO2 is a kind of non-contact physiological signal, which can reveal people's potential emotions. However, at this stage, the number of bands to generate StO2 is too many to achieve real-time emotion detection. Therefore, based on the publicly available HSI stress database, we used a Linear Prediction (LP) algorithm to select only 8 characteristic bands from the original 106 bands to generate StO2 and performed the task of identifying psychological stress and physical stress. The experimental results showed that the recognition rate of StO2 generated based on the selected 8 bands in stress detection is very close to (even higher) that of the original 106 bands, and reaches 84.44% when using the Bayes classifier.
{"title":"StO2 Stress Detection Based on Fewer Wavelengths Through Linear Prediction Algorithm","authors":"Xiao Xiao, Xinyu Liu, Dairong Peng, Tong Chen","doi":"10.1109/ICCCS52626.2021.9449292","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449292","url":null,"abstract":"In recent years, tissue oxygen saturation (StO2) based on hyperspectral imaging (HSI) technology has been deeply studied in the field of affective computing. HSI StO2 is a kind of non-contact physiological signal, which can reveal people's potential emotions. However, at this stage, the number of bands to generate StO2 is too many to achieve real-time emotion detection. Therefore, based on the publicly available HSI stress database, we used a Linear Prediction (LP) algorithm to select only 8 characteristic bands from the original 106 bands to generate StO2 and performed the task of identifying psychological stress and physical stress. The experimental results showed that the recognition rate of StO2 generated based on the selected 8 bands in stress detection is very close to (even higher) that of the original 106 bands, and reaches 84.44% when using the Bayes classifier.","PeriodicalId":376290,"journal":{"name":"2021 IEEE 6th International Conference on Computer and Communication Systems (ICCCS)","volume":"7 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":"125230020","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/ICCCS52626.2021.9449228
Xinhua Liu, Lu Zhu, Shuchang Xu, Shunpeng Du
Palette-based image recoloring is a popular method in the field of color processing by Intuitively converting the main color of the image into a palette that the user can directly manipulate. However, the traditional palette-based image recoloring method for natural images under different illuminations is affected by the light, colors on the extracted palette may be too close, which will cause color overflow and serious distortion after recoloring. This paper proposes a palette-based recoloring algorithm based on intrinsic decomposition. By using an intrinsic decomposition system based on different illumination images for unsupervised deep learning, natural image is decomposed into reflectance image and shading image. Only the reflectance image is recolored, and then combined with the shading image to obtain a natural recolored image. At the same time, in order to balance accuracy and ease of operation, we use the gap statistic algorithm to get the palette size. The recoloring results of natural images show the superiority of this method in terms of color transfer quality and user experience.
{"title":"Palette-Based Recoloring of Natural Images Under Different Illumination","authors":"Xinhua Liu, Lu Zhu, Shuchang Xu, Shunpeng Du","doi":"10.1109/ICCCS52626.2021.9449228","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449228","url":null,"abstract":"Palette-based image recoloring is a popular method in the field of color processing by Intuitively converting the main color of the image into a palette that the user can directly manipulate. However, the traditional palette-based image recoloring method for natural images under different illuminations is affected by the light, colors on the extracted palette may be too close, which will cause color overflow and serious distortion after recoloring. This paper proposes a palette-based recoloring algorithm based on intrinsic decomposition. By using an intrinsic decomposition system based on different illumination images for unsupervised deep learning, natural image is decomposed into reflectance image and shading image. Only the reflectance image is recolored, and then combined with the shading image to obtain a natural recolored image. At the same time, in order to balance accuracy and ease of operation, we use the gap statistic algorithm to get the palette size. The recoloring results of natural images show the superiority of this method in terms of color transfer quality and user experience.","PeriodicalId":376290,"journal":{"name":"2021 IEEE 6th International Conference on Computer and Communication Systems (ICCCS)","volume":"68 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":"131689991","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/ICCCS52626.2021.9449139
Nie Chao, Xu Chao, Feng Bo, Chi Yue
Specular reflections always exist in endoscopic images, and they can severely disturb surgeons' observation and judgment. This paper proposes a exemplar-based inpainting algorithm based on the adaptive search range to remove the endoscopic specular reflections. First, a binary thresholding algorithm based on b-channel and morphological dilation operation is used to automatically locate specular reflection regions, and then an improved algorithm based on the Criminisi algorithm is proposed, which uses the method of sequentially repairing highlight regions to reduce the edge contour priority calculation times, and then adopt an adaptive search range strategy to reduce the amount of calculation in the process of searching for the best matching patch, while searching in the additional search area provided by the best frame to improve the accuracy of matching. Experimental results demonstrate that the proposed method can automatically and correctly not only locate but also remove specular reflection regions in endoscopic images. In addition, because the improved method takes much less time than existing methods, and the PSNR and SSIM are higher than existing methods, the specular reflections removal scheme in this paper is better than existing methods.
{"title":"Specular Reflections Removal for Endoscopic Images Based on Improved Criminisi Algorithm","authors":"Nie Chao, Xu Chao, Feng Bo, Chi Yue","doi":"10.1109/ICCCS52626.2021.9449139","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449139","url":null,"abstract":"Specular reflections always exist in endoscopic images, and they can severely disturb surgeons' observation and judgment. This paper proposes a exemplar-based inpainting algorithm based on the adaptive search range to remove the endoscopic specular reflections. First, a binary thresholding algorithm based on b-channel and morphological dilation operation is used to automatically locate specular reflection regions, and then an improved algorithm based on the Criminisi algorithm is proposed, which uses the method of sequentially repairing highlight regions to reduce the edge contour priority calculation times, and then adopt an adaptive search range strategy to reduce the amount of calculation in the process of searching for the best matching patch, while searching in the additional search area provided by the best frame to improve the accuracy of matching. Experimental results demonstrate that the proposed method can automatically and correctly not only locate but also remove specular reflection regions in endoscopic images. In addition, because the improved method takes much less time than existing methods, and the PSNR and SSIM are higher than existing methods, the specular reflections removal scheme in this paper is better than existing methods.","PeriodicalId":376290,"journal":{"name":"2021 IEEE 6th International Conference on Computer and Communication Systems (ICCCS)","volume":"38 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":"133207269","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/ICCCS52626.2021.9449302
Honghan Zhu, D. Liu, Jingyan Liu, Paul Liu, Hao Yin, Yulan Peng
It is a significant challenge to obtain accurate boundary of tumors due to much speckle noises. In this paper, we proposed some meaningful modules to address the problem. Firstly, a large amount of semantic information is needed to determine the boundary on account of fuzziness around the edge of breast tumor, we proposed residual multi-scale(RMS) block to collect larger receptive filed. Secondly, we redesigned the skip connections and combined Squeeze-and-Excitation(SE) block to incorporate information from different layers. Finally we introduced deep supervision and hybrid loss function to accelerate the convergence of network. Dice similarity coefficient and intersection over union(IoU) were used to evaluate segmentation results which were 94.69 % and 90.01 % respectively on test set. It is shown that our method is effective in this kind of problem.
由于散斑噪声较多，难以获得准确的肿瘤边界。本文提出了一些有意义的模块来解决这一问题。首先，由于乳腺肿瘤边缘的模糊性，需要大量的语义信息来确定边界，我们提出残差多尺度(RMS)块来收集更大的接受野;其次，我们重新设计了跳跃连接，并结合了挤压和激励(SE)块，以吸收来自不同层的信息。最后，我们引入了深度监督和混合损失函数来加速网络的收敛。采用Dice similarity coefficient和intersection over union(IoU)对分割结果进行评价，在测试集上的分割率分别为94.69%和90.01%。结果表明，该方法对这类问题是有效的。
{"title":"RMS-SE-UNet: A Segmentation Method for Tumors in Breast Ultrasound Images","authors":"Honghan Zhu, D. Liu, Jingyan Liu, Paul Liu, Hao Yin, Yulan Peng","doi":"10.1109/ICCCS52626.2021.9449302","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449302","url":null,"abstract":"It is a significant challenge to obtain accurate boundary of tumors due to much speckle noises. In this paper, we proposed some meaningful modules to address the problem. Firstly, a large amount of semantic information is needed to determine the boundary on account of fuzziness around the edge of breast tumor, we proposed residual multi-scale(RMS) block to collect larger receptive filed. Secondly, we redesigned the skip connections and combined Squeeze-and-Excitation(SE) block to incorporate information from different layers. Finally we introduced deep supervision and hybrid loss function to accelerate the convergence of network. Dice similarity coefficient and intersection over union(IoU) were used to evaluate segmentation results which were 94.69 % and 90.01 % respectively on test set. It is shown that our method is effective in this kind of problem.","PeriodicalId":376290,"journal":{"name":"2021 IEEE 6th International Conference on Computer and Communication Systems (ICCCS)","volume":"10 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":"132802556","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/ICCCS52626.2021.9449169
Min Xie, Weize Sun, Lei Huang, Chuanxiang Xu, Huochao Tan
The matrix completion problem, which recover the missing data from the observed ones, had been widely studied in recent years. Although deep learning techniques had been applied in varies fields, limited works had done on matrix recovery. In this paper, we proposed a new deep neural network (DNN) model by integrating optimization theory and deep learning technique to solve the matrix completion problem. A cascaded neural network that contains the idea of alternating optimization is trained, and the application of SAR data reconstruction and imaging is used for evaluation. Experimental results shown that the proposed model can achieve better performance with less computational complexity when the sampling rate is sufficiently low.
{"title":"Accurate and Efficient Matrix Completion Using Cascaded Deep Neural Network","authors":"Min Xie, Weize Sun, Lei Huang, Chuanxiang Xu, Huochao Tan","doi":"10.1109/ICCCS52626.2021.9449169","DOIUrl":"https://doi.org/10.1109/ICCCS52626.2021.9449169","url":null,"abstract":"The matrix completion problem, which recover the missing data from the observed ones, had been widely studied in recent years. Although deep learning techniques had been applied in varies fields, limited works had done on matrix recovery. In this paper, we proposed a new deep neural network (DNN) model by integrating optimization theory and deep learning technique to solve the matrix completion problem. A cascaded neural network that contains the idea of alternating optimization is trained, and the application of SAR data reconstruction and imaging is used for evaluation. Experimental results shown that the proposed model can achieve better performance with less computational complexity when the sampling rate is sufficiently low.","PeriodicalId":376290,"journal":{"name":"2021 IEEE 6th International Conference on Computer and Communication Systems (ICCCS)","volume":"32 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":"115343689","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：