The critical nature of satellite network traffic provides a challenging environment to detect intrusions. The intrusion detection method presented aims to raise an alert whenever satellite network signals begin to exh...The critical nature of satellite network traffic provides a challenging environment to detect intrusions. The intrusion detection method presented aims to raise an alert whenever satellite network signals begin to exhibit anomalous patterns determined by Euclidian distance metric. In line with anomaly-based intrusion detection systems, the method presented relies heavily on building a model of＂normal＂ through the creation of a signal dictionary using windowing and k-means clustering. The results of three signals fi＇om our case study are discussed to highlight the benefits and drawbacks of the method presented. Our preliminary results demonstrate that the clustering technique used has great potential for intrusion detection for non-periodic satellite network signals.展开更多
Seismic data interpolation,especially irregularly sampled data interpolation,is a critical task for seismic processing and subsequent interpretation.Recently,with the development of machine learning and deep learning,...Seismic data interpolation,especially irregularly sampled data interpolation,is a critical task for seismic processing and subsequent interpretation.Recently,with the development of machine learning and deep learning,convolutional neural networks(CNNs)are applied for interpolating irregularly sampled seismic data.CNN based approaches can address the apparent defects of traditional interpolation methods,such as the low computational efficiency and the difficulty on parameters selection.However,current CNN based methods only consider the temporal and spatial features of irregularly sampled seismic data,which fail to consider the frequency features of seismic data,i.e.,the multi-scale features.To overcome these drawbacks,we propose a wavelet-based convolutional block attention deep learning(W-CBADL)network for irregularly sampled seismic data reconstruction.We firstly introduce the discrete wavelet transform(DWT)and the inverse wavelet transform(IWT)to the commonly used U-Net by considering the multi-scale features of irregularly sampled seismic data.Moreover,we propose to adopt the convolutional block attention module(CBAM)to precisely restore sampled seismic traces,which could apply the attention to both channel and spatial dimensions.Finally,we adopt the proposed W-CBADL model to synthetic and pre-stack field data to evaluate its validity and effectiveness.The results demonstrate that the proposed W-CBADL model could reconstruct irregularly sampled seismic data more effectively and more efficiently than the state-of-the-art contrastive CNN based models.展开更多
In marine seismic exploration, ocean-bottom cable techniques accurately record the multicomponent seismic wavefield; however, the seismic wave propagation in fluid–solid media cannot be simulated by a single wave equ...In marine seismic exploration, ocean-bottom cable techniques accurately record the multicomponent seismic wavefield; however, the seismic wave propagation in fluid–solid media cannot be simulated by a single wave equation. In addition, when the seabed interface is irregular, traditional finite-difference schemes cannot simulate the seismic wave propagation across the irregular seabed interface. Therefore, an acoustic–elastic forward modeling and vector-based P-and S-wave separation method is proposed. In this method, we divide the fluid–solid elastic media with irregular interface into orthogonal grids and map the irregular interface in the Cartesian coordinates system into a horizontal interface in the curvilinear coordinates system of the computational domain using coordinates transformation. The acoustic and elastic wave equations in the curvilinear coordinates system are applied to the fluid and solid medium, respectively. At the irregular interface, the two equations are combined into an acoustic–elastic equation in the curvilinear coordinates system. We next introduce a full staggered-grid scheme to improve the stability of the numerical simulation. Thus, separate P-and S-wave equations in the curvilinear coordinates system are derived to realize the P-and S-wave separation method.展开更多
针对由于目标频繁遮挡、不规则运动导致的外观特征不可靠和运动特征难以获取的问题,提出一种基于膨胀交并比区域(dilatation intersection over union,DIOU)匹配和自适应轨迹管理策略的多目标跟踪算法。DIOU模块通过膨胀匹配区域,提升...针对由于目标频繁遮挡、不规则运动导致的外观特征不可靠和运动特征难以获取的问题,提出一种基于膨胀交并比区域(dilatation intersection over union,DIOU)匹配和自适应轨迹管理策略的多目标跟踪算法。DIOU模块通过膨胀匹配区域,提升轨迹级联匹配的精度。自适应轨迹管理策略利用目标检测置信度动态调整轨迹生命周期,显著减少了异常跟踪和身份跳变。在公开数据集MOT17、MOT20和DanceTrack上进行了验证与测试,其在测试集上的高阶跟踪精度平均提升了2.4%,实验结果证明了所提方法的有效性。展开更多
文摘The critical nature of satellite network traffic provides a challenging environment to detect intrusions. The intrusion detection method presented aims to raise an alert whenever satellite network signals begin to exhibit anomalous patterns determined by Euclidian distance metric. In line with anomaly-based intrusion detection systems, the method presented relies heavily on building a model of＂normal＂ through the creation of a signal dictionary using windowing and k-means clustering. The results of three signals fi＇om our case study are discussed to highlight the benefits and drawbacks of the method presented. Our preliminary results demonstrate that the clustering technique used has great potential for intrusion detection for non-periodic satellite network signals.
基金Supported by the National Natural Science Foundation of China under Grant 42274144 and under Grant 41974137.
文摘Seismic data interpolation,especially irregularly sampled data interpolation,is a critical task for seismic processing and subsequent interpretation.Recently,with the development of machine learning and deep learning,convolutional neural networks(CNNs)are applied for interpolating irregularly sampled seismic data.CNN based approaches can address the apparent defects of traditional interpolation methods,such as the low computational efficiency and the difficulty on parameters selection.However,current CNN based methods only consider the temporal and spatial features of irregularly sampled seismic data,which fail to consider the frequency features of seismic data,i.e.,the multi-scale features.To overcome these drawbacks,we propose a wavelet-based convolutional block attention deep learning(W-CBADL)network for irregularly sampled seismic data reconstruction.We firstly introduce the discrete wavelet transform(DWT)and the inverse wavelet transform(IWT)to the commonly used U-Net by considering the multi-scale features of irregularly sampled seismic data.Moreover,we propose to adopt the convolutional block attention module(CBAM)to precisely restore sampled seismic traces,which could apply the attention to both channel and spatial dimensions.Finally,we adopt the proposed W-CBADL model to synthetic and pre-stack field data to evaluate its validity and effectiveness.The results demonstrate that the proposed W-CBADL model could reconstruct irregularly sampled seismic data more effectively and more efficiently than the state-of-the-art contrastive CNN based models.
基金financially supported by the Natural Science Foundation of China(No.41774133)the Open Funds of SINOPEC Key Laboratory of Geophysics(No.wtyjy-wx2017-01-04)National Science and Technology Major Project of the Ministry of Science and Technology of China(No.2016ZX05024-003-011)
文摘In marine seismic exploration, ocean-bottom cable techniques accurately record the multicomponent seismic wavefield; however, the seismic wave propagation in fluid–solid media cannot be simulated by a single wave equation. In addition, when the seabed interface is irregular, traditional finite-difference schemes cannot simulate the seismic wave propagation across the irregular seabed interface. Therefore, an acoustic–elastic forward modeling and vector-based P-and S-wave separation method is proposed. In this method, we divide the fluid–solid elastic media with irregular interface into orthogonal grids and map the irregular interface in the Cartesian coordinates system into a horizontal interface in the curvilinear coordinates system of the computational domain using coordinates transformation. The acoustic and elastic wave equations in the curvilinear coordinates system are applied to the fluid and solid medium, respectively. At the irregular interface, the two equations are combined into an acoustic–elastic equation in the curvilinear coordinates system. We next introduce a full staggered-grid scheme to improve the stability of the numerical simulation. Thus, separate P-and S-wave equations in the curvilinear coordinates system are derived to realize the P-and S-wave separation method.
文摘针对由于目标频繁遮挡、不规则运动导致的外观特征不可靠和运动特征难以获取的问题,提出一种基于膨胀交并比区域(dilatation intersection over union,DIOU)匹配和自适应轨迹管理策略的多目标跟踪算法。DIOU模块通过膨胀匹配区域,提升轨迹级联匹配的精度。自适应轨迹管理策略利用目标检测置信度动态调整轨迹生命周期,显著减少了异常跟踪和身份跳变。在公开数据集MOT17、MOT20和DanceTrack上进行了验证与测试,其在测试集上的高阶跟踪精度平均提升了2.4%,实验结果证明了所提方法的有效性。
