Pub Date : 2021-01-24DOI: 10.23919/Eusipco47968.2020.9287511
Arthur Marmin, M. Castella, J. Pesquet
While global optimization is a challenging topic in the nonconvex setting, a recent approach for optimizing polynomials reformulates the problem as an equivalent problem on measures, which is called a moment problem. It is then relaxed into a convex semidefinite programming problem whose solution gives the first moments of a measure supporting the optimal points. However, extracting the global solutions to the polynomial problem from those moments is still difficult, especially if the latter are poorly estimated. In this paper, we address the issue of extracting optimal points and interpret it as a tensor decomposition problem. By leveraging tools developed for noisy tensor decomposition, we propose a method to find the global solutions to a polynomial optimization problem from a noisy estimation of the solution of its corresponding moment problem. Finally, the interest of tensor decomposition methods for global polynomial optimization is shown through a detailed case study.
{"title":"Globally Optimizing Owing to Tensor Decomposition","authors":"Arthur Marmin, M. Castella, J. Pesquet","doi":"10.23919/Eusipco47968.2020.9287511","DOIUrl":"https://doi.org/10.23919/Eusipco47968.2020.9287511","url":null,"abstract":"While global optimization is a challenging topic in the nonconvex setting, a recent approach for optimizing polynomials reformulates the problem as an equivalent problem on measures, which is called a moment problem. It is then relaxed into a convex semidefinite programming problem whose solution gives the first moments of a measure supporting the optimal points. However, extracting the global solutions to the polynomial problem from those moments is still difficult, especially if the latter are poorly estimated. In this paper, we address the issue of extracting optimal points and interpret it as a tensor decomposition problem. By leveraging tools developed for noisy tensor decomposition, we propose a method to find the global solutions to a polynomial optimization problem from a noisy estimation of the solution of its corresponding moment problem. Finally, the interest of tensor decomposition methods for global polynomial optimization is shown through a detailed case study.","PeriodicalId":6705,"journal":{"name":"2020 28th European Signal Processing Conference (EUSIPCO)","volume":"76 1","pages":"990-994"},"PeriodicalIF":0.0,"publicationDate":"2021-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73114228","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 : 2021-01-24DOI: 10.23919/Eusipco47968.2020.9287619
M. Zervou, E. Doutsi, P. Pavlidis, P. Tsakalides
Prediction of protein structural classes from amino acid sequences is a challenging problem as it is profitable for analyzing protein function, interactions, and regulation. The majority of existing prediction methods for low-homology sequences utilize numerous amount of features and require an exhausting search for optimal parameter tuning. To address this problem, this work proposes a novel self-tuned architecture for feature extraction by modeling directly the inherent dynamics of the data in higher-dimensional phase space via chaos game representation (CGR) and generalized multidimensional recurrence quantification analysis (GmdRQA). Experimental evaluation on a real benchmark dataset demonstrates the superiority of the herein proposed architecture when compared against the state-of-the-art unidimensional RQA taking under consideration that our method achieves similar performance in a data-driven manner with a smaller computational cost.
{"title":"Efficient Dynamic Analysis of Low-similarity Proteins for Structural Class Prediction","authors":"M. Zervou, E. Doutsi, P. Pavlidis, P. Tsakalides","doi":"10.23919/Eusipco47968.2020.9287619","DOIUrl":"https://doi.org/10.23919/Eusipco47968.2020.9287619","url":null,"abstract":"Prediction of protein structural classes from amino acid sequences is a challenging problem as it is profitable for analyzing protein function, interactions, and regulation. The majority of existing prediction methods for low-homology sequences utilize numerous amount of features and require an exhausting search for optimal parameter tuning. To address this problem, this work proposes a novel self-tuned architecture for feature extraction by modeling directly the inherent dynamics of the data in higher-dimensional phase space via chaos game representation (CGR) and generalized multidimensional recurrence quantification analysis (GmdRQA). Experimental evaluation on a real benchmark dataset demonstrates the superiority of the herein proposed architecture when compared against the state-of-the-art unidimensional RQA taking under consideration that our method achieves similar performance in a data-driven manner with a smaller computational cost.","PeriodicalId":6705,"journal":{"name":"2020 28th European Signal Processing Conference (EUSIPCO)","volume":"55 1","pages":"1328-1332"},"PeriodicalIF":0.0,"publicationDate":"2021-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74300495","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 : 2021-01-24DOI: 10.23919/Eusipco47968.2020.9287334
Yang You, T. Oechtering
Non-dispersive infrared gas sensing is one of the best gas measurement method for air quality monitoring. However, sensors drift over time due to sensor aging and environmental factors, which makes calibration necessary. In this paper, we propose a hidden Markov model approach for sensor self-calibration, which builds on the physical model of gas sensors based on the Beer-Lambert law. We focus on the statistical dependency between a calibration coefficient and the temperature change. Supervised and unsupervised learning algorithms to learn the stochastic parameters of the hidden Markov model are derived and numerically tested. The true calibration coefficient at each time instant is estimated using the Viterbi algorithm. The numerical experiments using CO2 sensor data show excellent initial results which confirms that data-driven calibration of non-dispersive infrared gas sensors is possible. Meanwhile, the challenge in the practical design is to find an appropriate quantization scheme to keep the computation burden reasonable while achieving good performance.
{"title":"Hidden Markov Model Based Data-driven Calibration of Non-dispersive Infrared Gas Sensor","authors":"Yang You, T. Oechtering","doi":"10.23919/Eusipco47968.2020.9287334","DOIUrl":"https://doi.org/10.23919/Eusipco47968.2020.9287334","url":null,"abstract":"Non-dispersive infrared gas sensing is one of the best gas measurement method for air quality monitoring. However, sensors drift over time due to sensor aging and environmental factors, which makes calibration necessary. In this paper, we propose a hidden Markov model approach for sensor self-calibration, which builds on the physical model of gas sensors based on the Beer-Lambert law. We focus on the statistical dependency between a calibration coefficient and the temperature change. Supervised and unsupervised learning algorithms to learn the stochastic parameters of the hidden Markov model are derived and numerically tested. The true calibration coefficient at each time instant is estimated using the Viterbi algorithm. The numerical experiments using CO2 sensor data show excellent initial results which confirms that data-driven calibration of non-dispersive infrared gas sensors is possible. Meanwhile, the challenge in the practical design is to find an appropriate quantization scheme to keep the computation burden reasonable while achieving good performance.","PeriodicalId":6705,"journal":{"name":"2020 28th European Signal Processing Conference (EUSIPCO)","volume":"10 1","pages":"1717-1721"},"PeriodicalIF":0.0,"publicationDate":"2021-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72671712","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 : 2021-01-24DOI: 10.23919/Eusipco47968.2020.9287620
K. Schiecke, F. Benninger, M. Feucht
Investigations into brain-heart interactions are gaining increasing importance in various fields of research including epilepsy. Convergent Cross Mapping (CCM) is one method to quantify such interactions and was adapted for the analysis of children with temporal lobe epilepsy (TLE) in the past. Increasing amount of data and data features available produce a high and still rising complexity of results of such interaction analyses. Therefore, aim of this study was the investigation of generalized presentation of those results using our benchmark data set of children with TLE. Tensor decomposition was adapted to take into account spatial, time, frequency, directional and focus side related modes of interactions results achieved by CCM analysis.
{"title":"Analysis of Brain-Heart Couplings in Epilepsy: Dealing With the Highly Complex Structure of Resulting Interaction Pattern","authors":"K. Schiecke, F. Benninger, M. Feucht","doi":"10.23919/Eusipco47968.2020.9287620","DOIUrl":"https://doi.org/10.23919/Eusipco47968.2020.9287620","url":null,"abstract":"Investigations into brain-heart interactions are gaining increasing importance in various fields of research including epilepsy. Convergent Cross Mapping (CCM) is one method to quantify such interactions and was adapted for the analysis of children with temporal lobe epilepsy (TLE) in the past. Increasing amount of data and data features available produce a high and still rising complexity of results of such interaction analyses. Therefore, aim of this study was the investigation of generalized presentation of those results using our benchmark data set of children with TLE. Tensor decomposition was adapted to take into account spatial, time, frequency, directional and focus side related modes of interactions results achieved by CCM analysis.","PeriodicalId":6705,"journal":{"name":"2020 28th European Signal Processing Conference (EUSIPCO)","volume":"56 1","pages":"935-939"},"PeriodicalIF":0.0,"publicationDate":"2021-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74091194","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 : 2021-01-24DOI: 10.23919/Eusipco47968.2020.9287327
David S. Johnson, S. Grollmisch
The field of Industrial Sound Analysis (ISA) aims to automatically identify faults in production machinery or manufactured goods by analyzing audio signals. Publications in this field have shown that the surface condition of metal balls and different types of bulk materials (screws, nuts, etc.) sliding down a tube can be classified with a high accuracy using audio signals and deep neural networks. However, these systems suffer from domain shift, or dataset bias, due to minor changes in the recording setup which may easily happen in real-world production lines. This paper aims at finding methods to increase robustness of existing detection systems to domain shift, ideally without the need to record new data or retrain the models. Through five experiments, we implement a convolutional neural network (CNN) for two publicly available ISA datasets and evaluate transfer learning, data normalization and data augmentation as approaches to deal with domain shift. Our results show that while supervised methods with additional labeled data are the best approach, an unsupervised method that implements data augmentation with adaptive normalization is able to improve the performance by a large margin without the need of retraining neural networks.
{"title":"Techniques Improving the Robustness of Deep Learning Models for Industrial Sound Analysis","authors":"David S. Johnson, S. Grollmisch","doi":"10.23919/Eusipco47968.2020.9287327","DOIUrl":"https://doi.org/10.23919/Eusipco47968.2020.9287327","url":null,"abstract":"The field of Industrial Sound Analysis (ISA) aims to automatically identify faults in production machinery or manufactured goods by analyzing audio signals. Publications in this field have shown that the surface condition of metal balls and different types of bulk materials (screws, nuts, etc.) sliding down a tube can be classified with a high accuracy using audio signals and deep neural networks. However, these systems suffer from domain shift, or dataset bias, due to minor changes in the recording setup which may easily happen in real-world production lines. This paper aims at finding methods to increase robustness of existing detection systems to domain shift, ideally without the need to record new data or retrain the models. Through five experiments, we implement a convolutional neural network (CNN) for two publicly available ISA datasets and evaluate transfer learning, data normalization and data augmentation as approaches to deal with domain shift. Our results show that while supervised methods with additional labeled data are the best approach, an unsupervised method that implements data augmentation with adaptive normalization is able to improve the performance by a large margin without the need of retraining neural networks.","PeriodicalId":6705,"journal":{"name":"2020 28th European Signal Processing Conference (EUSIPCO)","volume":"1 1","pages":"81-85"},"PeriodicalIF":0.0,"publicationDate":"2021-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84611561","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 : 2021-01-24DOI: 10.23919/Eusipco47968.2020.9287628
M. Almasri, A. Mansour, C. Moy, A. Assoum, D. L. Jeune, C. Osswald
This manuscript investigates the problem of the Multi-Armed Bandit (MAB) in the context of the Opportunistic Spectrum Access (OSA) case with priority management (e.g. military applications). The main aim of a Secondary User (SU) in OSA is to increase his transmission throughput by seeking the best channel with the highest vacancy probability. In this manuscript, we propose a novel MAB algorithm called ϵ -UCB in order to enhance the spectrum learning of a SU and decrease the regret, i.e. the loss of reward due to the selection of worst channels. We analytically prove, and corroborate with simulations, that the regret of the proposed algorithm has a logarithmic behavior. So, after a finite number of time slots, the SU can estimate the vacancy probability of channels in order to target the best one for transmitting. Hereinafter, we extend ϵ -UCB to consider multiple priority users, where a SU can selfishly estimate and access the channels according to his prior rank. The simulation results show the superiority of the proposed algorithm for a single or multi-user cases compared to the well-known MAB algorithms.
{"title":"Managing Single or Multi-Users Channel Allocation for the Priority Cognitive Access","authors":"M. Almasri, A. Mansour, C. Moy, A. Assoum, D. L. Jeune, C. Osswald","doi":"10.23919/Eusipco47968.2020.9287628","DOIUrl":"https://doi.org/10.23919/Eusipco47968.2020.9287628","url":null,"abstract":"This manuscript investigates the problem of the Multi-Armed Bandit (MAB) in the context of the Opportunistic Spectrum Access (OSA) case with priority management (e.g. military applications). The main aim of a Secondary User (SU) in OSA is to increase his transmission throughput by seeking the best channel with the highest vacancy probability. In this manuscript, we propose a novel MAB algorithm called ϵ -UCB in order to enhance the spectrum learning of a SU and decrease the regret, i.e. the loss of reward due to the selection of worst channels. We analytically prove, and corroborate with simulations, that the regret of the proposed algorithm has a logarithmic behavior. So, after a finite number of time slots, the SU can estimate the vacancy probability of channels in order to target the best one for transmitting. Hereinafter, we extend ϵ -UCB to consider multiple priority users, where a SU can selfishly estimate and access the channels according to his prior rank. The simulation results show the superiority of the proposed algorithm for a single or multi-user cases compared to the well-known MAB algorithms.","PeriodicalId":6705,"journal":{"name":"2020 28th European Signal Processing Conference (EUSIPCO)","volume":"22 1","pages":"1722-1726"},"PeriodicalIF":0.0,"publicationDate":"2021-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84941373","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 : 2021-01-24DOI: 10.23919/Eusipco47968.2020.9287741
Amlu Anna Joshy, R. Rajan
Dysarthria is a neuro-motor speech disorder that renders speech unintelligible, in proportional to its severity. Assessing the severity level of dysarthria, apart from being a diagnostic step to evaluate the patient's improvement, is also capable of aiding automatic dysarthric speech recognition systems. In this paper, a detailed study on dysarthia severity classification using various deep learning architectural choices, namely deep neural network (DNN), convolutional neural network (CNN) and long short-term memory network (LSTM) is carried out. Mel frequency cepstral coefficients (MFCCs) and its derivatives are used as features. Performance of these models are compared with a baseline support vector machine (SVM) classifier using the UA-Speech corpus and the TORGO database. The highest classification accuracy of 96.18% and 93.24% are reported for TORGO and UA-Speech respectively. Detailed analysis on performance of these models shows that a proper choice of a deep learning architecture can ensure better performance than the conventionally used SVM classifier.
{"title":"Automated Dysarthria Severity Classification Using Deep Learning Frameworks","authors":"Amlu Anna Joshy, R. Rajan","doi":"10.23919/Eusipco47968.2020.9287741","DOIUrl":"https://doi.org/10.23919/Eusipco47968.2020.9287741","url":null,"abstract":"Dysarthria is a neuro-motor speech disorder that renders speech unintelligible, in proportional to its severity. Assessing the severity level of dysarthria, apart from being a diagnostic step to evaluate the patient's improvement, is also capable of aiding automatic dysarthric speech recognition systems. In this paper, a detailed study on dysarthia severity classification using various deep learning architectural choices, namely deep neural network (DNN), convolutional neural network (CNN) and long short-term memory network (LSTM) is carried out. Mel frequency cepstral coefficients (MFCCs) and its derivatives are used as features. Performance of these models are compared with a baseline support vector machine (SVM) classifier using the UA-Speech corpus and the TORGO database. The highest classification accuracy of 96.18% and 93.24% are reported for TORGO and UA-Speech respectively. Detailed analysis on performance of these models shows that a proper choice of a deep learning architecture can ensure better performance than the conventionally used SVM classifier.","PeriodicalId":6705,"journal":{"name":"2020 28th European Signal Processing Conference (EUSIPCO)","volume":"1 1","pages":"116-120"},"PeriodicalIF":0.0,"publicationDate":"2021-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84127160","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 : 2021-01-24DOI: 10.23919/Eusipco47968.2020.9287220
Zhuyin Li, A. Giorgetti, S. Kandeepan
The unique features of unmanned aerial vehicles (UAVs) extend a large number of existing technologies into environments that are not suitable for on-site operations. Localization, a critical basis of many applications such as cognitive radio and first response networks, can benefit UAV technology as well. In such scenarios, an underinvestigated problem is the non-collaborative localization of multiple primary users (PUs). Therefore, this work proposes a data-driven multiple PU localization algorithm based on the angular and power measurements performed by a UAV equipped with an antenna array. The measured data firstly generate a score map, then a threshold and a hierarchical clustering method are applied to the score map to both detect the number of PUs and estimate their location. The performance of the algorithm is assessed by numerical results in terms of probability of detecting the number of PUs, and root-mean-square-error of position estimation. The proposed solution exhibit remarkable performance considering that the approach requires only the knowledge of the PUs frequency band.
{"title":"UAV Mapping for Multiple Primary Users Localization","authors":"Zhuyin Li, A. Giorgetti, S. Kandeepan","doi":"10.23919/Eusipco47968.2020.9287220","DOIUrl":"https://doi.org/10.23919/Eusipco47968.2020.9287220","url":null,"abstract":"The unique features of unmanned aerial vehicles (UAVs) extend a large number of existing technologies into environments that are not suitable for on-site operations. Localization, a critical basis of many applications such as cognitive radio and first response networks, can benefit UAV technology as well. In such scenarios, an underinvestigated problem is the non-collaborative localization of multiple primary users (PUs). Therefore, this work proposes a data-driven multiple PU localization algorithm based on the angular and power measurements performed by a UAV equipped with an antenna array. The measured data firstly generate a score map, then a threshold and a hierarchical clustering method are applied to the score map to both detect the number of PUs and estimate their location. The performance of the algorithm is assessed by numerical results in terms of probability of detecting the number of PUs, and root-mean-square-error of position estimation. The proposed solution exhibit remarkable performance considering that the approach requires only the knowledge of the PUs frequency band.","PeriodicalId":6705,"journal":{"name":"2020 28th European Signal Processing Conference (EUSIPCO)","volume":"7 1","pages":"1787-1791"},"PeriodicalIF":0.0,"publicationDate":"2021-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84141492","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 : 2021-01-24DOI: 10.23919/Eusipco47968.2020.9287684
João Silva, Marco Oliveira, Aníbal J. S. Ferreira
Whispered-voice to normal-voice conversion is typically achieved using codec-based analysis and re-synthesis, using statistical conversion of important spectral and prosodic features, or using data-driven end-to-end signal conversion. These approaches are however highly constrained by the architecture of the codec, the statistical projection, or the size and quality of the training data. In this paper, we presume direct implantation of voiced phonemes in whispered speech and we focus on fully flexible parametric models that i) can be independently controlled, ii) synthesize natural and linguistically correct voiced phonemes, iii) preserve idiosyncratic characteristics of a given speaker, and iv) are amenable to co-articulation effects through simple model interpolation. We use natural spoken and sung vowels to illustrate these capabilities in a signal modeling and re-synthesis process where spectral magnitude, phase structure, F0 contour and sound morphing can be independently controlled in arbitrary ways.
{"title":"Flexible parametric implantation of voicing in whispered speech under scarce training data","authors":"João Silva, Marco Oliveira, Aníbal J. S. Ferreira","doi":"10.23919/Eusipco47968.2020.9287684","DOIUrl":"https://doi.org/10.23919/Eusipco47968.2020.9287684","url":null,"abstract":"Whispered-voice to normal-voice conversion is typically achieved using codec-based analysis and re-synthesis, using statistical conversion of important spectral and prosodic features, or using data-driven end-to-end signal conversion. These approaches are however highly constrained by the architecture of the codec, the statistical projection, or the size and quality of the training data. In this paper, we presume direct implantation of voiced phonemes in whispered speech and we focus on fully flexible parametric models that i) can be independently controlled, ii) synthesize natural and linguistically correct voiced phonemes, iii) preserve idiosyncratic characteristics of a given speaker, and iv) are amenable to co-articulation effects through simple model interpolation. We use natural spoken and sung vowels to illustrate these capabilities in a signal modeling and re-synthesis process where spectral magnitude, phase structure, F0 contour and sound morphing can be independently controlled in arbitrary ways.","PeriodicalId":6705,"journal":{"name":"2020 28th European Signal Processing Conference (EUSIPCO)","volume":"39 1","pages":"416-420"},"PeriodicalIF":0.0,"publicationDate":"2021-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84535073","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：