Pub Date : 2024-09-17DOI: 10.1109/LSP.2024.3462174
Xulong Liu;Weixiang Li;Kaiqing Lin;Bin Li
Current deep learning-based JPEG image steganalysis methods typically rely on decompressed pixels for steganalytic feature extraction, without fully leveraging the inherent information in JPEG images. Additionally, they often face limitations such as large parameter counts and restricted image sizes for detection. In this letter, we propose a spatial-frequency feature fusion network (SF3Net) for lightweight and arbitrary-sized JPEG steganalysis. SF3Net introduces a PReLU activation function and a multi-view convolutional module to capture refined residual features from decompressed pixels, while also integrating original DCT coefficients and quantization tables to extract additional modal features. The spatial-frequency multi-modality features are then fused using a coordinate attention mechanism. And a patch splitting scheme is designed to be compatible with any feature resolution, enabling the detection of arbitrary-sized images with a Swin Transformer block. Experimental results demonstrate that SF3Net outperforms existing methods in detecting both fixed-sized and arbitrary-sized images, while significantly reducing the number of parameters.
{"title":"Spatial-Frequency Feature Fusion Network for Lightweight and Arbitrary-Sized JPEG Steganalysis","authors":"Xulong Liu;Weixiang Li;Kaiqing Lin;Bin Li","doi":"10.1109/LSP.2024.3462174","DOIUrl":"10.1109/LSP.2024.3462174","url":null,"abstract":"Current deep learning-based JPEG image steganalysis methods typically rely on decompressed pixels for steganalytic feature extraction, without fully leveraging the inherent information in JPEG images. Additionally, they often face limitations such as large parameter counts and restricted image sizes for detection. In this letter, we propose a spatial-frequency feature fusion network (SF3Net) for lightweight and arbitrary-sized JPEG steganalysis. SF3Net introduces a PReLU activation function and a multi-view convolutional module to capture refined residual features from decompressed pixels, while also integrating original DCT coefficients and quantization tables to extract additional modal features. The spatial-frequency multi-modality features are then fused using a coordinate attention mechanism. And a patch splitting scheme is designed to be compatible with any feature resolution, enabling the detection of arbitrary-sized images with a Swin Transformer block. Experimental results demonstrate that SF3Net outperforms existing methods in detecting both fixed-sized and arbitrary-sized images, while significantly reducing the number of parameters.","PeriodicalId":13154,"journal":{"name":"IEEE Signal Processing Letters","volume":"31 ","pages":"2585-2589"},"PeriodicalIF":3.2,"publicationDate":"2024-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142260632","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-17DOI: 10.1109/LSP.2024.3462292
Taoshen Li;Shuping Dang;Zhihui Ge;Zhenrong Zhang
Because of the equilibrium between mathematical tractability and approximation accuracy maintained by the inverse Gaussian (IG) distributional model, it has been regarded as the most appropriate approximation substitute for the lognormal distributional model for shadowed and atmospheric turbulence induced (ATI) fading in the past decades. In this paper, we conduct an in-depth information-theoretic analysis for the lognormal-to-IG channel model substitution (CMS) technique and study its parametric mapping optimality achieved by minimizing the Kullback-Leibler (K-L) divergence between the two distributional models. In this way, we rigorously prove that the moment matching criterion produces the optimal IG substitute for lognormal reference distributions, which has never been observed in other CMS techniques. In addition, we clarify a myth in the realm of CMS that the IG substitute outperforms the gamma substitute for approximating lognormal reference distributions; instead, the substitution superiority shall depend on the parametric mapping criterion and the scale parameter of the lognormal reference distribution. All analytical insights presented in this paper are validated by simulation results.
{"title":"On the Optimality of Inverse Gaussian Approximation for Lognormal Channel Models","authors":"Taoshen Li;Shuping Dang;Zhihui Ge;Zhenrong Zhang","doi":"10.1109/LSP.2024.3462292","DOIUrl":"10.1109/LSP.2024.3462292","url":null,"abstract":"Because of the equilibrium between mathematical tractability and approximation accuracy maintained by the inverse Gaussian (IG) distributional model, it has been regarded as the most appropriate approximation substitute for the lognormal distributional model for shadowed and atmospheric turbulence induced (ATI) fading in the past decades. In this paper, we conduct an in-depth information-theoretic analysis for the lognormal-to-IG channel model substitution (CMS) technique and study its parametric mapping optimality achieved by minimizing the Kullback-Leibler (K-L) divergence between the two distributional models. In this way, we rigorously prove that the moment matching criterion produces the optimal IG substitute for lognormal reference distributions, which has never been observed in other CMS techniques. In addition, we clarify a myth in the realm of CMS that the IG substitute outperforms the gamma substitute for approximating lognormal reference distributions; instead, the substitution superiority shall depend on the parametric mapping criterion and the scale parameter of the lognormal reference distribution. All analytical insights presented in this paper are validated by simulation results.","PeriodicalId":13154,"journal":{"name":"IEEE Signal Processing Letters","volume":"31 ","pages":"2625-2629"},"PeriodicalIF":3.2,"publicationDate":"2024-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142260635","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-16DOI: 10.1109/LSP.2024.3461654
Zhuoran Zheng;Chen Wu;Yeying Jin;Xiuyi Jia
Recently, large models (Segment Anything model) came on the scene to provide a new baseline for polyp segmentation tasks. This demonstrates that large models with a sufficient image level prior can achieve promising performance on a given task. In this paper, we unfold a new perspective on polyp segmentation modeling by leveraging the Depth Anything Model (DAM) to provide depth prior to polyp segmentation models. Specifically, the input polyp image is first passed through a frozen DAM to generate a depth map. The depth map and the input polyp images are then concatenated and fed into a convolutional neural network with multiscale to generate segmented images. Extensive experimental results demonstrate the effectiveness of our method, and in addition, we observe that our method still performs well on images of polyps with noise.
最近，大型模型（Segment Anything model）的出现为息肉分割任务提供了新的基准。这表明，具有足够图像级先验的大型模型可以在给定任务中取得可喜的性能。在本文中，我们利用深度任意模型（DAM）为息肉分割模型提供深度先验，从而为息肉分割建模提供了一个新的视角。具体来说，输入的息肉图像首先通过冻结的 DAM 生成深度图。然后将深度图和输入的息肉图像连接起来，并输入具有多尺度的卷积神经网络，生成分割图像。广泛的实验结果证明了我们方法的有效性，此外，我们还观察到我们的方法在有噪声的息肉图像上仍然表现良好。
{"title":"Polyp-DAM: Polyp Segmentation via Depth Anything Model","authors":"Zhuoran Zheng;Chen Wu;Yeying Jin;Xiuyi Jia","doi":"10.1109/LSP.2024.3461654","DOIUrl":"https://doi.org/10.1109/LSP.2024.3461654","url":null,"abstract":"Recently, large models (Segment Anything model) came on the scene to provide a new baseline for polyp segmentation tasks. This demonstrates that large models with a sufficient image level prior can achieve promising performance on a given task. In this paper, we unfold a new perspective on polyp segmentation modeling by leveraging the Depth Anything Model (DAM) to provide depth prior to polyp segmentation models. Specifically, the input polyp image is first passed through a frozen DAM to generate a depth map. The depth map and the input polyp images are then concatenated and fed into a convolutional neural network with multiscale to generate segmented images. Extensive experimental results demonstrate the effectiveness of our method, and in addition, we observe that our method still performs well on images of polyps with noise.","PeriodicalId":13154,"journal":{"name":"IEEE Signal Processing Letters","volume":"31 ","pages":"2925-2929"},"PeriodicalIF":3.2,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142524188","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Radar signal deinterleaving is an essential step in perceiving the battlefield situation and mastering military initiative in the information battlefield. Complex radar systems are rapidly updated and iterated, which exacerbates the possibility of “increasing batch” and “mistaken batch” during radar signal deinterleaving. In this letter, a novel method based on complex networks and Laplacian graph clustering is proposed to improve the accuracy of deinterleaving. First, a complex network is constructed to mine the spatial correlation relationships of the same radar signals. Then, based on the graph characteristics of the Laplacian matrix, the number of cluster centers is solved. Finally, this letter employs Laplacian spectral clustering based on graph segmentation to accomplish radar signal deinterleaving. The results of the experimental simulation demonstrate that the method is capable of effectively tackling the “increasing batch” and “mistaken batch” problems of radar signal deinterleaving, and could reach 99.88% deinterleaving accuracy with high robustness.
{"title":"A Radar Signal Deinterleaving Method Based on Complex Network and Laplacian Graph Clustering","authors":"Qiang Guo;Shuai Huang;Liangang Qi;Daren Li;Mykola Kaliuzhnyi","doi":"10.1109/LSP.2024.3461656","DOIUrl":"https://doi.org/10.1109/LSP.2024.3461656","url":null,"abstract":"Radar signal deinterleaving is an essential step in perceiving the battlefield situation and mastering military initiative in the information battlefield. Complex radar systems are rapidly updated and iterated, which exacerbates the possibility of “increasing batch” and “mistaken batch” during radar signal deinterleaving. In this letter, a novel method based on complex networks and Laplacian graph clustering is proposed to improve the accuracy of deinterleaving. First, a complex network is constructed to mine the spatial correlation relationships of the same radar signals. Then, based on the graph characteristics of the Laplacian matrix, the number of cluster centers is solved. Finally, this letter employs Laplacian spectral clustering based on graph segmentation to accomplish radar signal deinterleaving. The results of the experimental simulation demonstrate that the method is capable of effectively tackling the “increasing batch” and “mistaken batch” problems of radar signal deinterleaving, and could reach 99.88% deinterleaving accuracy with high robustness.","PeriodicalId":13154,"journal":{"name":"IEEE Signal Processing Letters","volume":"31 ","pages":"2580-2584"},"PeriodicalIF":3.2,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142359734","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-16DOI: 10.1109/LSP.2024.3459811
Natsuki Akaishi;Koki Yamada;Kohei Yatabe
Harmonic/percussive source separation (HPSS) is an important tool for analyzing and processing audio signals. The standard approach to HPSS takes advantage of the structural difference of sinusoidal and percussive components, called anisotropic smoothness , in magnitude spectrograms. However, the existing methods disregard phase of the spectrograms and/or approximate the problem, which naturally limits the upper bound of the performance of HPSS. In this letter, we propose a novel approach to HPSS that regards phase without the approximation. The proposed method introduces an auxiliary variable that acts as an adaptive weight of a weighted energy minimization problem, which enables us to apply smoothing on magnitude of complex-valued spectrograms. Compared to the existing methods, the proposed method can obtain separated components having better magnitude and phase by simultaneously handling them.
{"title":"Harmonic/Percussive Source Separation Based on Anisotropic Smoothness of Magnitude Spectrograms via Convex Optimization","authors":"Natsuki Akaishi;Koki Yamada;Kohei Yatabe","doi":"10.1109/LSP.2024.3459811","DOIUrl":"https://doi.org/10.1109/LSP.2024.3459811","url":null,"abstract":"Harmonic/percussive source separation (HPSS) is an important tool for analyzing and processing audio signals. The standard approach to HPSS takes advantage of the structural difference of sinusoidal and percussive components, called \u0000<italic>anisotropic smoothness</i>\u0000, in magnitude spectrograms. However, the existing methods disregard phase of the spectrograms and/or approximate the problem, which naturally limits the upper bound of the performance of HPSS. In this letter, we propose a novel approach to HPSS that regards phase without the approximation. The proposed method introduces an auxiliary variable that acts as an adaptive weight of a weighted energy minimization problem, which enables us to apply smoothing on magnitude of complex-valued spectrograms. Compared to the existing methods, the proposed method can obtain separated components having better magnitude and phase by simultaneously handling them.","PeriodicalId":13154,"journal":{"name":"IEEE Signal Processing Letters","volume":"31 ","pages":"2575-2579"},"PeriodicalIF":3.2,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142359735","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The acquisition of large-scale paired visible and thermal images is crucial for enhancing face recognition systems, especially in low-light environments where visible spectrum images fail. However, the task is hindered by the scarcity of thermal images and the need for identity consistency during image generation. In this paper, we propose Pair-ID, an innovative framework that addresses these challenges by creating a shared latent space for simultaneous generation of paired visible and thermal images. Pair-ID integrates identity information into text embeddings and employs fixed templates for diverse facial poses, streamlining the customization process and reducing computational demands. The framework's Joint Learner encodes both modalities, facilitating synchronized image generation and preserving facial details. Extensive evaluations show that Pair-ID surpasses current methods in efficiency and performance for paired data generation, making it a promising solution for face recognition under varying lighting conditions.
{"title":"Pair-ID: A Dual Modal Framework for Identity Preserving Image Generation","authors":"Jingyu Lin;Yongrong Wu;Zeyu Wang;Xiaode Liu;Yufei Guo","doi":"10.1109/LSP.2024.3461648","DOIUrl":"10.1109/LSP.2024.3461648","url":null,"abstract":"The acquisition of large-scale paired visible and thermal images is crucial for enhancing face recognition systems, especially in low-light environments where visible spectrum images fail. However, the task is hindered by the scarcity of thermal images and the need for identity consistency during image generation. In this paper, we propose Pair-ID, an innovative framework that addresses these challenges by creating a shared latent space for simultaneous generation of paired visible and thermal images. Pair-ID integrates identity information into text embeddings and employs fixed templates for diverse facial poses, streamlining the customization process and reducing computational demands. The framework's Joint Learner encodes both modalities, facilitating synchronized image generation and preserving facial details. Extensive evaluations show that Pair-ID surpasses current methods in efficiency and performance for paired data generation, making it a promising solution for face recognition under varying lighting conditions.","PeriodicalId":13154,"journal":{"name":"IEEE Signal Processing Letters","volume":"31 ","pages":"2715-2719"},"PeriodicalIF":3.2,"publicationDate":"2024-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142260636","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-13DOI: 10.1109/LSP.2024.3460478
Jun Lu;Lei Liu;Shunqi Huang;Ning Wei;Xiaoming Chen
Approximate message passing (AMP) algorithms are iterative methods for signal recovery in noisy linear systems. In some scenarios, AMP algorithms need to operate within a distributed network. To address this challenge, the distributed extensions of AMP (D-AMP, FD-AMP) and orthogonal/vector AMP (D-OAMP/D-VAMP) were proposed, but they still inherit the limitations of centralized algorithms. In this letter, we propose distributed memory AMP (D-MAMP) to overcome the IID matrix limitation of D-AMP/FD-AMP, as well as the high complexity and heavy communication cost of D-OAMP/D-VAMP. We introduce a matrix-by-vector variant of MAMP tailored for distributed computing. Leveraging this variant, D-MAMP enables each node to execute computations utilizing locally available observation vectors and transform matrices. Meanwhile, global summations of locally updated results are conducted through message interaction among nodes. For acyclic graphs, D-MAMP converges to the same mean square error performance as the centralized MAMP.
{"title":"Distributed Memory Approximate Message Passing","authors":"Jun Lu;Lei Liu;Shunqi Huang;Ning Wei;Xiaoming Chen","doi":"10.1109/LSP.2024.3460478","DOIUrl":"10.1109/LSP.2024.3460478","url":null,"abstract":"Approximate message passing (AMP) algorithms are iterative methods for signal recovery in noisy linear systems. In some scenarios, AMP algorithms need to operate within a distributed network. To address this challenge, the distributed extensions of AMP (D-AMP, FD-AMP) and orthogonal/vector AMP (D-OAMP/D-VAMP) were proposed, but they still inherit the limitations of centralized algorithms. In this letter, we propose distributed memory AMP (D-MAMP) to overcome the IID matrix limitation of D-AMP/FD-AMP, as well as the high complexity and heavy communication cost of D-OAMP/D-VAMP. We introduce a matrix-by-vector variant of MAMP tailored for distributed computing. Leveraging this variant, D-MAMP enables each node to execute computations utilizing locally available observation vectors and transform matrices. Meanwhile, global summations of locally updated results are conducted through message interaction among nodes. For acyclic graphs, D-MAMP converges to the same mean square error performance as the centralized MAMP.","PeriodicalId":13154,"journal":{"name":"IEEE Signal Processing Letters","volume":"31 ","pages":"2660-2664"},"PeriodicalIF":3.2,"publicationDate":"2024-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142260637","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-13DOI: 10.1109/LSP.2024.3460483
Tong Zhan;Xiaojun Mao;Jian Wang;Zhonglei Wang
Collective matrix completion (CMC) offers a straightforward approach to dealing with data with entries from various sources. Benefiting from the joint structure in the collective matrix, CMC often achieves fast convergence. However, since CMC conducts matrix-level operations, it neglects the entry-wise information that can potentially be very useful for matrix completion. In this paper, to capture the entry-wise information, we propose a method called graph collective matrix completion (GCoMC). Specifically, our method integrates a graph pattern extraction module into CMC via a relational graph convolutional network. Experiments on simulated and real-world datasets show that our method significantly outperforms some existing counterparts.
{"title":"Collective Matrix Completion via Graph Extraction","authors":"Tong Zhan;Xiaojun Mao;Jian Wang;Zhonglei Wang","doi":"10.1109/LSP.2024.3460483","DOIUrl":"10.1109/LSP.2024.3460483","url":null,"abstract":"Collective matrix completion (CMC) offers a straightforward approach to dealing with data with entries from various sources. Benefiting from the joint structure in the collective matrix, CMC often achieves fast convergence. However, since CMC conducts matrix-level operations, it neglects the entry-wise information that can potentially be very useful for matrix completion. In this paper, to capture the entry-wise information, we propose a method called graph collective matrix completion (GCoMC). Specifically, our method integrates a graph pattern extraction module into CMC via a relational graph convolutional network. Experiments on simulated and real-world datasets show that our method significantly outperforms some existing counterparts.","PeriodicalId":13154,"journal":{"name":"IEEE Signal Processing Letters","volume":"31 ","pages":"2620-2624"},"PeriodicalIF":3.2,"publicationDate":"2024-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142269614","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-12DOI: 10.1109/LSP.2024.3460654
Zhi Chai;Xinyong Peng;Xinran Huang;Mingye Li;Xuelin Yang
Robust radio frequency fingerprint identification (RFFI) is crucial for physical layer authentication, while it suffers from channel effects and requires extra overhead to increase recognition accuracy (RA). To address this, an efficient channel-robust RFFI scheme is proposed, employing a specialized multi-task learning (MTL) framework to direct the neural network (NN) toward extracting channel-robust features. In addition, receiver collaboration (RC) is utilized for data augmentation and output calibration. Experimental results demonstrate that the RA is significantly increased from 51.72% to 99.97% when using the open-resource Wi-Fi signal datasets collected from different time periods. Meanwhile, the requirements for extra data transmission, NN structure, and feature crafting in the inferring stage are dramatically simplified.
{"title":"Channel-Robust RF Fingerprint Identification Using Multi-Task Learning and Receiver Collaboration","authors":"Zhi Chai;Xinyong Peng;Xinran Huang;Mingye Li;Xuelin Yang","doi":"10.1109/LSP.2024.3460654","DOIUrl":"10.1109/LSP.2024.3460654","url":null,"abstract":"Robust radio frequency fingerprint identification (RFFI) is crucial for physical layer authentication, while it suffers from channel effects and requires extra overhead to increase recognition accuracy (RA). To address this, an efficient channel-robust RFFI scheme is proposed, employing a specialized multi-task learning (MTL) framework to direct the neural network (NN) toward extracting channel-robust features. In addition, receiver collaboration (RC) is utilized for data augmentation and output calibration. Experimental results demonstrate that the RA is significantly increased from 51.72% to 99.97% when using the open-resource Wi-Fi signal datasets collected from different time periods. Meanwhile, the requirements for extra data transmission, NN structure, and feature crafting in the inferring stage are dramatically simplified.","PeriodicalId":13154,"journal":{"name":"IEEE Signal Processing Letters","volume":"31 ","pages":"2510-2514"},"PeriodicalIF":3.2,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142218746","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","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：