Pub Date : 2018-11-01DOI: 10.1109/GlobalSIP.2018.8646500
Xueyuan Wang, M. C. Gursoy
In this paper, we consider simultaneous wireless information and power transfer in millimeter wave (mmWave) cellular networks with user-centric base station deployments. The distinguishing features of mmWave communications are incorporated into the system model. Moreover, the locations of user equipments (UEs) are modeled as a Thomas cluster process. First, the association probability is investigated. Subsequently, using tools from stochastic geometry, we analyze the energy coverage and signal-to-interference-plus-noise ratio (SINR) coverage of the network and provide general expressions. Through numerical results, we draw insights on how to model the system to improve the coverage performance.
{"title":"JOINT ENERGY AND SINR COVERAGE IN ENERGY HARVESTING MMWAVE CELLULAR NETWORKS WITH USER-CENTRIC BASE STATION DEPLOYMENTS","authors":"Xueyuan Wang, M. C. Gursoy","doi":"10.1109/GlobalSIP.2018.8646500","DOIUrl":"https://doi.org/10.1109/GlobalSIP.2018.8646500","url":null,"abstract":"In this paper, we consider simultaneous wireless information and power transfer in millimeter wave (mmWave) cellular networks with user-centric base station deployments. The distinguishing features of mmWave communications are incorporated into the system model. Moreover, the locations of user equipments (UEs) are modeled as a Thomas cluster process. First, the association probability is investigated. Subsequently, using tools from stochastic geometry, we analyze the energy coverage and signal-to-interference-plus-noise ratio (SINR) coverage of the network and provide general expressions. Through numerical results, we draw insights on how to model the system to improve the coverage performance.","PeriodicalId":119131,"journal":{"name":"2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129801543","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 : 2018-11-01DOI: 10.1109/GLOBALSIP.2018.8646534
Harry Sevi, G. Rilling, P. Borgnat
In this paper, we discuss the problem of modeling a graph signal on a directed graph when observing only partially the graph signal. The graph signal is recovered using a learned graph filter. The novelty is to use the random walk operator associated to an ergodic random walk on the graph, so as to define and learn a graph filter, expressed as a polynomial of this operator. Through the study of different cases, we show the efficiency of the signal modeling using the random walk operator compared to existing methods using the adjacency matrix or ignoring the directions in the graph.
{"title":"MODELING SIGNALS OVER DIRECTED GRAPHS THROUGH FILTERING","authors":"Harry Sevi, G. Rilling, P. Borgnat","doi":"10.1109/GLOBALSIP.2018.8646534","DOIUrl":"https://doi.org/10.1109/GLOBALSIP.2018.8646534","url":null,"abstract":"In this paper, we discuss the problem of modeling a graph signal on a directed graph when observing only partially the graph signal. The graph signal is recovered using a learned graph filter. The novelty is to use the random walk operator associated to an ergodic random walk on the graph, so as to define and learn a graph filter, expressed as a polynomial of this operator. Through the study of different cases, we show the efficiency of the signal modeling using the random walk operator compared to existing methods using the adjacency matrix or ignoring the directions in the graph.","PeriodicalId":119131,"journal":{"name":"2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129345142","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 : 2018-11-01DOI: 10.1109/GlobalSIP.2018.8646378
Ekhi Uranga, Á. Llorente, A. D. L. Fuente
The European Space Agency’s Soil Moisture and Ocean Salinity (SMOS) mission operates in the 1400-1427 MHz frequency band, which is allocated to the EESS (passive) service in the ITU Radio-Regulations. The measurements of SMOS radiometer are perturbed by radio frequency interference (RFI) that jeopardize part of its scientific retrieval in certain areas of the World.The strategies initiated by the European Space Agency to mitigate the impact of RFI includes the detection, monitoring, and reporting of the interference cases. Due to the large number of sources detected, their temporal variability, and the fluid contacts with some National Administrations, it was necessary to automate the RFI mitigation process.This paper presents the database created for the classification of the RFI sources and their details, including a website for queries and reports using the stored data. In addition, the algorithms developed to automate the detections that populate the database are explained.
{"title":"DATABASE OF SMOS RFI SOURCES IN THE 1400-1427MHZ PASSIVE BAND","authors":"Ekhi Uranga, Á. Llorente, A. D. L. Fuente","doi":"10.1109/GlobalSIP.2018.8646378","DOIUrl":"https://doi.org/10.1109/GlobalSIP.2018.8646378","url":null,"abstract":"The European Space Agency’s Soil Moisture and Ocean Salinity (SMOS) mission operates in the 1400-1427 MHz frequency band, which is allocated to the EESS (passive) service in the ITU Radio-Regulations. The measurements of SMOS radiometer are perturbed by radio frequency interference (RFI) that jeopardize part of its scientific retrieval in certain areas of the World.The strategies initiated by the European Space Agency to mitigate the impact of RFI includes the detection, monitoring, and reporting of the interference cases. Due to the large number of sources detected, their temporal variability, and the fluid contacts with some National Administrations, it was necessary to automate the RFI mitigation process.This paper presents the database created for the classification of the RFI sources and their details, including a website for queries and reports using the stored data. In addition, the algorithms developed to automate the detections that populate the database are explained.","PeriodicalId":119131,"journal":{"name":"2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP)","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124642042","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 : 2018-11-01DOI: 10.1109/GlobalSIP.2018.8646521
O. A. Zoubi, Ahmad Mayeli, V. Zotev, H. Refai, M. Paulus, J. Bodurka
Electroencephalography (EEG) has been widely used in human brain research. Several techniques in EEG relies on analyzing the topographical distribution of the data. One of the most common analysis is EEG microstates (EEG-ms). EEG-ms reflects the stable topographical representation of EEG signal lasting a few dozen milliseconds. EEG-ms were associated with resting state fMRI networks and related mental processes and abnormalities. One challenge in EEG-ms analysis is the polarity invariant property for the signal, in which the relative direction of local minima and maxima is taking into consideration. Thus, identifying those topographies requires special handling for the data using modified clustering algorithms. Here, we propose a polarity invariant transformation for EEG data to eliminate the difficulties with handling the polarity of the data during the EEG-ms identification part, which would allow better clustering EEG data. Our results demonstrate how the transformation work and show the benefit of using such a transformation.
{"title":"POLARITY INVARIANT TRANSFORMATION FOR EEG MICROSTATES ANALYSIS","authors":"O. A. Zoubi, Ahmad Mayeli, V. Zotev, H. Refai, M. Paulus, J. Bodurka","doi":"10.1109/GlobalSIP.2018.8646521","DOIUrl":"https://doi.org/10.1109/GlobalSIP.2018.8646521","url":null,"abstract":"Electroencephalography (EEG) has been widely used in human brain research. Several techniques in EEG relies on analyzing the topographical distribution of the data. One of the most common analysis is EEG microstates (EEG-ms). EEG-ms reflects the stable topographical representation of EEG signal lasting a few dozen milliseconds. EEG-ms were associated with resting state fMRI networks and related mental processes and abnormalities. One challenge in EEG-ms analysis is the polarity invariant property for the signal, in which the relative direction of local minima and maxima is taking into consideration. Thus, identifying those topographies requires special handling for the data using modified clustering algorithms. Here, we propose a polarity invariant transformation for EEG data to eliminate the difficulties with handling the polarity of the data during the EEG-ms identification part, which would allow better clustering EEG data. Our results demonstrate how the transformation work and show the benefit of using such a transformation.","PeriodicalId":119131,"journal":{"name":"2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP)","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131473106","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 proliferation of 3D sensors induced 3D computer vision research for many application areas including virtual reality, autonomous navigation and surveillance. Recently, different methods have been proposed for 3D object classification. Many of the existing 2D and 3D classification methods rely on convolutional neural networks (CNNs), which are very successful in extracting features from the data. However, CNNs cannot sufficiently address the spatial relationship between features due to the max-pooling layers, and they require vast amount of training data. In this paper, we propose a model architecture for 3D object classification, which is an extension of Capsule Networks (CapsNets) to 3D data. Our proposed architecture called 3D CapsNet, takes advantage of the fact that a CapsNet preserves the orientation and spatial relationship of the extracted features, and thus requires less data to train the network. We compare our approach with ShapeNet on the ModelNet database, and show that our method provides performance improvement especially when training data size gets smaller.
{"title":"Object Classification from 3D Volumetric Data with 3D Capsule Networks","authors":"Burak Kakillioglu, Ayesha Ahmad, Senem Velipasalar","doi":"10.1109/GlobalSIP.2018.8646333","DOIUrl":"https://doi.org/10.1109/GlobalSIP.2018.8646333","url":null,"abstract":"The proliferation of 3D sensors induced 3D computer vision research for many application areas including virtual reality, autonomous navigation and surveillance. Recently, different methods have been proposed for 3D object classification. Many of the existing 2D and 3D classification methods rely on convolutional neural networks (CNNs), which are very successful in extracting features from the data. However, CNNs cannot sufficiently address the spatial relationship between features due to the max-pooling layers, and they require vast amount of training data. In this paper, we propose a model architecture for 3D object classification, which is an extension of Capsule Networks (CapsNets) to 3D data. Our proposed architecture called 3D CapsNet, takes advantage of the fact that a CapsNet preserves the orientation and spatial relationship of the extracted features, and thus requires less data to train the network. We compare our approach with ShapeNet on the ModelNet database, and show that our method provides performance improvement especially when training data size gets smaller.","PeriodicalId":119131,"journal":{"name":"2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP)","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116226960","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 : 2018-11-01DOI: 10.1109/GlobalSIP.2018.8646634
Hanwook Chung, É. Plourde, B. Champagne
In this paper, we introduce a supervised multi-channel speech enhancement algorithm based on a Bayesian multi-channel non-negative matrix factorization (MNMF) model. In the proposed framework, we consider the probabilistic generative model (PGM) of MNMF, specified by Poisson-distributed latent variables and gamma-distributed priors. In the training stage, the MNMF parameters of the speech and noise sources are estimated via the variational Bayesian expectation-maximization (VBEM) algorithm. In the enhancement stage, the clean speech signal is estimated via the MNMF-based minimum variance distortionless response (MVDR) beamformer. To further improve the enhanced speech quality, we efficiently combine the MNMF-based beamforming technique with a classical unsupervised single-channel enhancement method. Experiments show that the proposed method can provide better enhancement performance than the selected benchmarks.
{"title":"A SUPERVISED MULTI-CHANNEL SPEECH ENHANCEMENT ALGORITHM BASED ON BAYESIAN NMF MODEL","authors":"Hanwook Chung, É. Plourde, B. Champagne","doi":"10.1109/GlobalSIP.2018.8646634","DOIUrl":"https://doi.org/10.1109/GlobalSIP.2018.8646634","url":null,"abstract":"In this paper, we introduce a supervised multi-channel speech enhancement algorithm based on a Bayesian multi-channel non-negative matrix factorization (MNMF) model. In the proposed framework, we consider the probabilistic generative model (PGM) of MNMF, specified by Poisson-distributed latent variables and gamma-distributed priors. In the training stage, the MNMF parameters of the speech and noise sources are estimated via the variational Bayesian expectation-maximization (VBEM) algorithm. In the enhancement stage, the clean speech signal is estimated via the MNMF-based minimum variance distortionless response (MVDR) beamformer. To further improve the enhanced speech quality, we efficiently combine the MNMF-based beamforming technique with a classical unsupervised single-channel enhancement method. Experiments show that the proposed method can provide better enhancement performance than the selected benchmarks.","PeriodicalId":119131,"journal":{"name":"2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP)","volume":" 7","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120833232","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 : 2018-11-01DOI: 10.1109/GlobalSIP.2018.8646330
Trang C. Mai, H. Ngo, T. Duong
In this paper, we investigate the impact of multiple-antenna deployment at access points (APs) and users on the performance of cell-free massive multiple-input multiple-output (MIMO). The transmission is done via time-division duplex (TDD) protocol. With this protocol, the channels are first estimated at each AP based on the received pilot signals in the training phase. Then these channel information will be used to decode the symbols before sending to all users. The simple and distributed conjugate beamforming technique is deployed. We derive a closed-form expression for the downlink spectral efficiency taking into account the imperfect channel state information (CSI), non-orthogonal pilots, and power control. This spectral efficiency can be achieved without the knowledge of instantaneous CSI at the users. In addition, the effects of the number antennas per APs and per users are analyzed in the case of using mutual orthogonal pilot sequences and data power control.
{"title":"CELL-FREE MASSIVE MIMO SYSTEMS WITH MULTI-ANTENNA USERS","authors":"Trang C. Mai, H. Ngo, T. Duong","doi":"10.1109/GlobalSIP.2018.8646330","DOIUrl":"https://doi.org/10.1109/GlobalSIP.2018.8646330","url":null,"abstract":"In this paper, we investigate the impact of multiple-antenna deployment at access points (APs) and users on the performance of cell-free massive multiple-input multiple-output (MIMO). The transmission is done via time-division duplex (TDD) protocol. With this protocol, the channels are first estimated at each AP based on the received pilot signals in the training phase. Then these channel information will be used to decode the symbols before sending to all users. The simple and distributed conjugate beamforming technique is deployed. We derive a closed-form expression for the downlink spectral efficiency taking into account the imperfect channel state information (CSI), non-orthogonal pilots, and power control. This spectral efficiency can be achieved without the knowledge of instantaneous CSI at the users. In addition, the effects of the number antennas per APs and per users are analyzed in the case of using mutual orthogonal pilot sequences and data power control.","PeriodicalId":119131,"journal":{"name":"2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP)","volume":"220 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121036824","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 : 2018-11-01DOI: 10.1109/GLOBALSIP.2018.8646651
Pu Zhao, Kaidi Xu, Tianyun Zhang, M. Fardad, Yanzhi Wang, X. Lin
As deep learning penetrates into wide application domains, it is essential to evaluate the robustness of deep neural networks (DNNs) under adversarial attacks, especially for some security-critical applications. To better understand the security properties of DNNs, we propose a general framework for constructing adversarial examples, based on ADMM (Alternating Direction Method of Multipliers). This general framework can be adapted to implement L2 and L0 attacks with minor changes. Our ADMM attacks require less distortion for incorrect classification compared with C&W attacks. Our ADMM attack is also able to break defenses such as defensive distillation and adversarial training, and provide strong attack transferability.
{"title":"Reinforced Adversarial Attacks on Deep Neural Networks Using ADMM","authors":"Pu Zhao, Kaidi Xu, Tianyun Zhang, M. Fardad, Yanzhi Wang, X. Lin","doi":"10.1109/GLOBALSIP.2018.8646651","DOIUrl":"https://doi.org/10.1109/GLOBALSIP.2018.8646651","url":null,"abstract":"As deep learning penetrates into wide application domains, it is essential to evaluate the robustness of deep neural networks (DNNs) under adversarial attacks, especially for some security-critical applications. To better understand the security properties of DNNs, we propose a general framework for constructing adversarial examples, based on ADMM (Alternating Direction Method of Multipliers). This general framework can be adapted to implement L2 and L0 attacks with minor changes. Our ADMM attacks require less distortion for incorrect classification compared with C&W attacks. Our ADMM attack is also able to break defenses such as defensive distillation and adversarial training, and provide strong attack transferability.","PeriodicalId":119131,"journal":{"name":"2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121094676","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 : 2018-11-01DOI: 10.1109/GlobalSIP.2018.8646367
Yantao Lu, Senem Velipasalar
Many methods have been proposed for human activity classification, which rely either on Inertial Measurement Unit (IMU) data or data from static cameras watching subjects. There have been relatively less work using egocentric videos, and even fewer approaches combining egocentric video and IMU data. Systems relying only on IMU data are limited in the complexity of the activities that they can detect. In this paper, we present a robust and autonomous method, for fine-grained activity classification, that leverages data from multiple wearable sensor modalities to differentiate between activities, which are similar in nature, with a level of accuracy that would be impossible by each sensor alone. We use both egocentric videos and IMU sensors on the body. We employ Capsule Networks together with Convolutional Long Short Term Memory (LSTM) to analyze egocentric videos, and an LSTM framework to analyze IMU data, and capture temporal aspect of actions. We performed experiments on the CMU-MMAC dataset achieving overall recall and precision rates of 85.8% and 86.2%, respectively. We also present results of using each sensor modality alone, which show that the proposed approach provides 19.47% and 39.34% increase in accuracy compared to using only ego-vision data and only IMU data, respectively.
{"title":"HUMAN ACTIVITY CLASSIFICATION INCORPORATING EGOCENTRIC VIDEO AND INERTIAL MEASUREMENT UNIT DATA","authors":"Yantao Lu, Senem Velipasalar","doi":"10.1109/GlobalSIP.2018.8646367","DOIUrl":"https://doi.org/10.1109/GlobalSIP.2018.8646367","url":null,"abstract":"Many methods have been proposed for human activity classification, which rely either on Inertial Measurement Unit (IMU) data or data from static cameras watching subjects. There have been relatively less work using egocentric videos, and even fewer approaches combining egocentric video and IMU data. Systems relying only on IMU data are limited in the complexity of the activities that they can detect. In this paper, we present a robust and autonomous method, for fine-grained activity classification, that leverages data from multiple wearable sensor modalities to differentiate between activities, which are similar in nature, with a level of accuracy that would be impossible by each sensor alone. We use both egocentric videos and IMU sensors on the body. We employ Capsule Networks together with Convolutional Long Short Term Memory (LSTM) to analyze egocentric videos, and an LSTM framework to analyze IMU data, and capture temporal aspect of actions. We performed experiments on the CMU-MMAC dataset achieving overall recall and precision rates of 85.8% and 86.2%, respectively. We also present results of using each sensor modality alone, which show that the proposed approach provides 19.47% and 39.34% increase in accuracy compared to using only ego-vision data and only IMU data, respectively.","PeriodicalId":119131,"journal":{"name":"2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP)","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127115042","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 : 2018-11-01DOI: 10.1109/globalsip.2018.8646458
{"title":"GlobalSIP 2018 Committees","authors":"","doi":"10.1109/globalsip.2018.8646458","DOIUrl":"https://doi.org/10.1109/globalsip.2018.8646458","url":null,"abstract":"","PeriodicalId":119131,"journal":{"name":"2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125915028","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：