检索结果-内蒙古大学图书馆

Leveraging neighborhood distance awareness for entity alignment in temporal knowledge graphs

NEURAL networkS 2025年 185卷 107181页

作者： Zhu, Lin Li, Guishun Bai, Luyi Northeastern Univ Qinhuangdao Hebei Key Lab Marine Percept Network & Data Proc Qinhuangdao 066004 Peoples R China

Entity alignment (EA) is a typical strategy for knowledge graph integration, aiming to identify and align different entity pairs representing the same real object from different knowledge graphs. Temporal Knowledge Graph (TKG) extends the static knowledge graph by introducing timestamps. However, since temporal knowledge graphs are constructed based on their own data sources, this usually leads to problems such as missing or redundant entity information in the temporal knowledge graph. Therefore, temporal knowledge graph entity alignment is a meaningful task to avoid the above problems. However, most existing methods tend to ignore the impact of direct and indirect neighborhood information on EA. Therefore, we propose a temporal knowledge graph entity alignment method with neighborhood distance awareness, namely Tem-DA. Tem-DA models direct and indirect neighbors separately, capturing directly connected neighbors with related entities and indirectly connected neighbors with unrelated entities through a distance detection module. Furthermore, we propose a gating mechanism weight the elements in the embedding matrix and through cross-entropy loss function with regularization terms to address adaptive fusion. This mechanism dynamically adjusts the weights of different feature embeddings, thus providing a more flexible and adaptive feature fusion strategy compared with traditional linear weight adjustment methods. Tem-DA effectively captures the temporal information of entities and take advantage of overlapping properties that may have multiple identical time intervals between different entities to encode the temporal information. For some entities that may lack sufficient temporal information, Tem-DA estimates the temporal characteristics of adjacent information and generates temporal embedding vectors. Experimental results on two monolingual TKG datasets show that Tem-DA outperforms popular baseline methods.

关键词： Entity alignment Gating mechanism Graph attention network Temporal knowledge graph

来源：评论

学校读者我要写书评

暂无评论

Visual perception enhancement fall detection algorithm based on vision transformer

引用

SIGNAL IMAGE AND VIDEO procESSING 2025年第1期19卷 1-12页

作者： Cai, Xi Wang, Xiangcheng Bao, Kexin Chen, Yinuo Jiao, Yin Han, Guang Northeastern Univ Qinhuangdao Sch Comp & Commun Engn Hebei Key Lab Marine Percept Network & Data Proc Qinhuangdao 066004 Peoples R China

Fall detection is a crucial research topic in public healthcare. With advances in intelligent surveillance and deep learning, vision-based fall detection has gained significant attention. While numerous deep learning algorithms prevail in video fall detection due to excellent feature processing capabilities, they all exhibit limitations in handling long-term spatiotemporal dependencies. Recently, Vision Transformer has shown considerable potential in integrating global information and understanding long-term spatiotemporal dependencies, thus providing novel solutions. In view of this, we propose a visual perception enhancement fall detection algorithm based on Vision Transformer. We utilize Vision Transformer-Base as the baseline model for analyzing global motion information in videos. On this basis, to address the model's difficulty in capturing subtle motion changes across video frames, we design an inter-frame motion information enhancement module. Concurrently, we propose a locality perception enhancement self-attention mechanism to overcome the model's weak focus on local key features within the frame. Experimental results show that our method achieves notable performance on the Le2i and UR datasets, surpassing several advanced methods.

关键词： Fall detection Computer vision Deep learning Vision transformer Attention mechanism

来源：评论

学校读者我要写书评

暂无评论

An adaptive Kalman filter loosely coupled indoor fusion positioning system based on inertial navigation system and ultra-wide band

引用

MEASUREMENT 2025年 244卷

作者： Cheng, Long Fu, Zhentao Northeastern Univ Dept Comp & Commun Engn Qinhuangdao 066004 Hebei Peoples R China Northeastern Univ Hebei Key Lab Marine Percept Network & Data Proc Qinhuangdao 066004 Hebei Peoples R China

Inertial navigation system (INS) indoor position technology is receiving increasing attention from researchers, especially the combination of inertial navigation and Ultra-wideband (UWB), which can achieve good results in indoor positioning. This article further explores and improves the positioning accuracy based on this model. This article proposes an optimization approach to address the issues of loose coupling in traditional inertial navigation, utilizing deep learning and improved particle filter to improve its accuracy. Firstly, we use an adaptive threshold gait detection algorithm and replaced traditional extended Kalman filter (EKF) with adaptive extended Kalman filter. Secondly, in terms of UWB data processing, compensation is provided for non-line-of-sight (NLOS) UWB data. Finally, the correct measurement values obtained from inertial navigation are fed into the particle filter improved by the firefly algorithm (FAPF) for final positioning. At the same time, we can note that the traditional use of inertial navigation to calculate the heading and step length to guide the motion of each particle in the particle filter has a significant problem, which is that the cumulative error is becoming increasingly large. Therefore, this article uses recurrent neural network (RNN) to regress three-axis acceleration and three-axis angular velocity to obtain step length and heading, in order to guide the motion of each particle. Through the final experiment, we obtained a positioning error of around 0.4 m.

关键词： Indoor localization Non-line-of-sight Inertial navigation system Firefly algorithm Kalman filter

来源：评论

学校读者我要写书评

暂无评论

An INS/UWB joint indoor positioning algorithm based on hypothesis testing and yaw angle

引用

WIRELESS networkS 2025年第1期31卷 457-469页

作者： Cheng, Long Zhao, Fuyang Zhao, Wenhao Northeastern Univ Dept Comp & Commun Engn Qinhuangdao 066004 Hebei Peoples R China Northeastern Univ Hebei Key Lab Marine Percept Network & Data Proc Qinhuangdao 066004 Hebei Peoples R China

Wireless sensor network (WSN) is widely used in indoor positioning, but indoor positioning is susceptible to non-line-of-sight (NLOS) propagation environment. The inertial navigation system (INS) does not depend on external information, but it will produce a large cumulative error when working for a long time. The combination of Ultra-wide band (UWB) positioning and inertial navigation positioning can not only effectively reduce the impact of NLOS interference, but also alleviate the impact of INS cumulative error. This paper proposes an algorithm based on yaw angle and UWB joint positioning. In order to weaken the cumulative error of the INS itself, this paper uses the UWB positioning results to correct the INS positioning data and yaw angle data through the extended Kalman filter (EKF), and then performs subsequent positioning according to the modified yaw angle until the next data correction. In addition, this algorithm uses a hypothesis test method for INS and UWB data processing, which weakens the error impact of environmental factors. The proposed algorithm is compared with existing algorithms using mean square error (RMSE) as an indicator. The simulation and experimental results show that the algorithm has better performance in NLOS interference environment.

关键词： Hypothesis testing Indoor position Inertial navigation system Ultra-wide band Yaw angle

来源：评论

学校读者我要写书评

暂无评论

Mc-music algorithm for fast DOA estimation

引用

WIRELESS networkS 2025年第1期31卷 659-667页

作者： Liu, Fulai Li, Xubin Su, Zhibo Liu, Yubiao Zhang, Aiyi Northeastern Univ Qinhuangdao Lab Cognit Radio & Big Spectrum Data Proc Hebei Key Lab Marine Percept Network & Data Proc Qinhuangdao 066004 Peoples R China Northeastern Univ Sch Comp Sci & Engn Shenyang 110819 Peoples R China

This paper presents a fast and compressible multiple signal classification algorithm (C-MUSIC) based on the manifold separation technique (MST) for a uniform circular array (UCA), named as MC-MUSIC. Making use of the MST, the steering vector (SV) of the UCA can be converted into a new SV, which has a Vandermonde structure similar to a uniform linear array. Then a noise-like subspace cluster (NLSC) can be generated by applying a new linear relation between the true direction-of-arrival (DOA) and virtual DOA to the new given SV. Finally, the angular field-of-view can be compressed to a small sector by the intersection of the NLSC so that the calculation times of the spatial spectrum can be reduced effectively. Simulations demonstrate the presented method can maintain the high precision of the DOA estimation with low computational complexity.

关键词： Uniform circular array Manifold separation technique Direction-of-arrival Multiple signal classification

来源：评论

学校读者我要写书评

暂无评论

YGNet: A Lightweight Object Detection Model for Remote Sensing

引用

IEEE GEOSCIENCE AND REMOTE SENSING LETTERS 2025年 22卷

作者： Song, Xin Gao, Erhao Northeastern Univ Sch Comp Sci & Engn Shenyang 110169 Peoples R China Northeastern Univ Hebei Key Lab Marine Percept Network & Data Proc Qinhuangdao 066004 Hebei Peoples R China

In the dynamic field of remote sensing images (RSIs), the challenge of object scale variability and sensor resolution disparities is formidable. Addressing these complexities, we have designed a lightweight remote sensing model named YGNet, tailored for multiscale object detection. It demonstrates excellent performance in detecting both multiscale and small objects within RSIs. The E-RMSK module within YGNet employs a gradient-based architecture with multiple parallel reparameterized convolutions in its internal branches, facilitating the extraction of multiscale features while maintaining parameter and computational efficiency. The HLS-PAN structure integrates feature maps extracted through feature selection, enabling the top layers to relay image information downward to lower levels and the lowest layers to transmit data upward for localization, achieving feature fusion. This synergistic effect of the module design enhances the accuracy of object detection in complex remote sensing scenarios and ensures the model's feasibility on platforms with limited resources. Rigorous testing on the RSOD and NWPU VHR-10 datasets has proven YGNet's exceptional capabilities, achieving the mean average precision (mAP) scores of 96.2% and 88.9%, respectively. The model meets the demands for real-time, lightweight, multiscale object detection in remote sensing imagery, making it highly suitable for deployment in resource-constrained environments.

关键词： Feature extraction Object detection Remote sensing Kernel Accuracy Semantics Computational modeling Neck Location awareness Computer architecture Lightweight model multiscale fusion object detection remote sensing

来源：评论

学校读者我要写书评

暂无评论

Doa estimation algorithm based on UV decomposition matrix completion

引用

WIRELESS networkS 2025年第3期31卷 2317-2326页

作者： Liu, Fulai Meng, Guangyu Zhang, Bo Zhang, Aiyi Lou, Xinyue Du, Ruiyan Northeastern Univ Qinhuangdao Lab Cognit Radio & Big Spectrum Data Proc Hebei Key Lab Marine Percept Network & Data Proc Qinhuangdao 066004 Peoples R China Northeastern Univ Sch Comp Sci & Engn Shenyang 110819 Peoples R China

In the imperfect communication environment, the array received signal may have missing data, which could suffer from the direction-of-arrival (DOA) estimation performance deterioration. To this end, the matrix completion method is usually used for recovering the missing data of received signal to improve the DOA estimation accuracy. In this paper, a matrix completion DOA estimation algorithm is proposed based on UV decomposition. Firstly, the array received signal recovery problem is transformed into a matrix completion problem. On this basis, the UV decomposition of the signal is introduced into the matrix completion model, which avoids the singular value decomposition of kernel norm optimization and reduces the computational complexity. In addition, a residual regular term is utilized in the proposed algorithm to reduce the data deviation caused by matrix decomposition. Then, the received signal of the array to be recovered is solved by the alternating direction multiplier method (ADMM) algorithm. Finally, the multiple signal classification (MUSIC) algorithm is used to estimate the DOA of recovered array signal. Simulation results indicate that the proposed algorithm can better recover data to perform DOA estimation with the low computational complexity, and is superior to existing DOA estimation algorithms based on matrix completion.

关键词： Array antenna Direction-of-arrival estimation Matrix completion UV decomposition

来源：评论

学校读者我要写书评

暂无评论

DSRSN algorithm for channel estimation in IRS assisted mmWave MIMO systems

引用

SIGNAL IMAGE AND VIDEO procESSING 2025年第6期19卷 1-7页

作者： Liu, Fulai Ge, Junjie Feng, Tian Du, Ruiyan Shi, Huiyang Northeastern Univ Qinhuangdao Hebei Key Lab Marine Percept Network & Data Proc Lab Cognit Radio & Big Spectrum Data Proc Qinhuangdao 066004 Peoples R China Northeastern Univ Sch Comp Sci & Engn Shenyang 110819 Peoples R China

The complexity of multi-dimensional channels makes accurate channel estimation difficult in intelligent reflecting surface (IRS) assisted mmWave MIMO systems. To improve the estimation accuracy, this paper presents a channel estimation algorithm based on a depthwise separable residual shrinkage network (DSRSN), named as DSRSN algorithm. Firstly, the channel matrix estimated by the least square method and the real channel matrix are given as the training input and the training label, separately. Then, a depthwise separable residual shrinkage frame is constructed to improve the accuracy of channel matrix estimation. Specially, to decrease the network parameters number, depthwise separable convolution is introduced. Next, the input features are processed by a soft threshold module to suppress channel noise and extract channel characteristics more accurately. Finally, the well-trained DSRSN model can exactly output the estimated channel matrix. According to the simulation results, the suggested DSRSN approach increases the estimation accuracy further while requiring less pilot overhead. For example, the simulation results verify that the normalized mean square error reduces by more than 2.94 times in comparision to the related methods, when the signal-to-noise ratio is 15dB.

关键词： Channel estimation Intelligent reflecting surface Depthwise separable convolution Residual shrinkage network

来源：评论

学校读者我要写书评

暂无评论

Multiple classification algorithm based on ensemble learning for intrusion detection

引用

WIRELESS networkS 2025年第3期31卷 2143-2154页

作者： Liu, Fulai Yue, Jiaqi Hu, Zhongyi Du, Ruiyan Northeastern Univ Qinhuangdao Hebei Key Lab Marine Percept Network & Data Proc Lab Cognit Radio & Big Spectrum Data Proc Qinhuangdao 066004 Peoples R China Northeastern Univ Sch Comp Sci & Engn Shenyang 110819 Peoples R China

As one of the promising technologies for enhancing network security, network intrusion detection is crucial for accurately identifying various network attacks. To advance the intrusion detection capability, a multiple classification ensemble learning algorithm (MCELID) is proposed by analyzing intrusion data from diverse perspectives. Firstly, an individual learner combining graph convolutional networks and long short-term memory networks (GCN-LSTM) is constructed to extract both structural information and temporal correlations from intrusion data. Meanwhile, a weighted support vector machine (W-SVM) model is employed for handling multi-classification tasks, where the conditional probability is calculated to derive the probability vector of input samples. Finally, the intrusion detection result is determined by a soft voting mechanism that combines the classification results from the GCN-LSTM and W-SVM models. Extensive experiments conducted on the KDD-CUP 99 dataset demonstrate that the proposed MCELID algorithm outperforms other existing methods in terms of detection accuracy, particularly in the recognition of DOS and Probe attacks.

关键词： Ensemble learning Intrusion detection Long short term memory Graph convolution neural network Support vector machine

来源：评论

学校读者我要写书评

暂无评论

Sparrow search algorithm optimized LSTM model for CSI-based indoor localization

引用

EARTH SCIENCE INFORMATICS 2025年第1期18卷 1-16页

作者： Wang, Yan Zhou, Yuqing Northeastern Univ Dept Comp & Commun Engn Qinhuangdao 066004 Hebei Peoples R China Northeastern Univ Hebei Key Lab Marine Percept Network & Data Proc Qinhuangdao 066004 Hebei Peoples R China

In indoor positioning, Channel State Information (CSI) holds significant importance. CSI provides detailed channel characteristics, including amplitude and phase information, enabling the positioning system to estimate device locations more accurately. Compared to traditional RSSI methods, CSI offers higher positioning accuracy and better interference resistance. Additionally, it supports various advanced positioning algorithms, making it suitable for complex indoor environments and enabling real-time precise positioning. In this study, a localization method based on CSI phase and amplitude data is proposed with the aim of improving the accuracy and robustness of indoor localization. In the offline training phase, for each reference point, we construct a new dataset using pre-processed amplitude and phase data, which are transformed into image data for pre-training to extract features. Subsequently, a CNN-LSTM neural network with an integrated attention mechanism is trained to capture complex spatio-temporal relationships in the data. However, manual tuning of neural network hyperparameters has its limitations. Therefore, this paper introduces the Sparrow Search Algorithm (SSA)bio-inspired algorithm to optimize the combination of hyperparameters, aiming to obtain a more optimal model performance. During the online orientation phase, CSI data collected from the target device is processed using the same steps as in the offline phase to convert it into images. The trained model is then used to perform regression predictions on test points, enabling localization of the target position. To assess the effectiveness of our proposed approach, we ran tests in both open and complicated laboratory settings. Comparisons were made with several existing indoor localization solutions (such as PCNB, MIMO, LSTM, and FIFS), demonstrating that our approach exhibits significant advantages regarding localization accuracy and robustness.

关键词： Indoor localization Channel state information Long short-term memory Sparrow search algorithm Attention mechanism

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：