检索结果-内蒙古大学图书馆

Sound event localization and detection based on deep learning

Journal of Systems Engineering and Electronics 2024年第2期35卷 294-301页

作者： ZHAO Dada DING Kai QI Xiaogang CHEN Yu FENG Hailin School of Mathematics and Statistics Xidian UniversityXi’an 710071China Science and Technology on Near-Surface Detection Laboratory Wuxi 214035China

Acoustic source localization(ASL)and sound event detection(SED)are two widely pursued independent research *** recent years,in order to achieve a more complete spatial and temporal representation of sound field,sound event localization and detection(SELD)has become a very active research *** paper presents a deep learning-based multioverlapping sound event localization and detection algorithm in three-dimensional ***-Mel spectrum and generalized cross-correlation spectrum are joined together in channel dimension as input *** features are classified and regressed in parallel after training by a neural network to obtain sound recognition and localization results *** channel attention mechanism is also introduced in the network to selectively enhance the features containing essential information and suppress the useless ***,a thourough comparison confirms the efficiency and effectiveness of the proposed SELD *** experiments show that the proposed algorithm is robust to reverberation and environment and can achieve higher recognition and localization accuracy compared with the baseline method.

关键词： sound event localization and detection(SELD) deep learning convolutional recursive neural network(CRNN) channel attention mechanism

来源：评论

学校读者我要写书评

暂无评论

TFANet: Vehicle Sound-based Classification Utilizing Time-Frequency Hybrid Attention

Journal of Network Intelligence

引用

Journal of Network Intelligence 2025年第1期10卷 112-129页

作者： Jiang, Bo Li, Hong-Bin Xu, Yue-Lin Ding, Kai Zhao, Xu-Dong Science and Technology on Near-Surface Detection Laboratory Wuxi214035 China Jiangsu Puxu Software Information Technology Co. Ltd Nanjing210001 China

Vehicle sound-based classification is essential for preserving the natural environment and intelligent transportation, helping monitor device operation, maintain environmental safety, and regulate vehicle traffic. Nevertheless, the intricate nature of outdoor situations, combined with elements like wind noise, brings substantial extraneous or disruptive information into the data, which has a detrimental impact on classification performance. To tackle this issue, this study introduces a vehicle classification network called the Time-Frequency Hybrid Attention Mechanism Classification Network(TFANet). Firstly this network is built on the mature ResNet18 architecture and utilizes Mel Frequency Cepstral Coefficients (MFCC) features to analyze sound signals, significantly improving the accuracy and robustness of the classification model. In addition, it designs a time-frequency hybrid attention mechanism to assign more importance to the semantic relevant time frames and key frequency bands within the MFCCs spectrogram, helping to minimize the impact of irrelevant information and ensure the accuracy of vehicle classification. Finally, we collected and organized the Field Vehicle Sound Dataset (FVSD), and conducted relevant experiments on this dataset, demonstrating that the approach achieved an F1 score of 0.950. Meanwhile, additional experiments were performed using the IDMT-Traffic open-access dataset. The TFANet showed an improved F1 score in each category compared to the baseline model of this dataset. Notably, the Truck category, which had the poorest classification performance, significantly improved by 0.16 in the F1 score. © 2025, Taiwan Ubiquitous Information CO LTD. All rights reserved.

关键词： Hybrid vehicles

来源：评论

学校读者我要写书评

暂无评论

Camouflaged people detection based on a semi-supervised search identification network

引用

Defence technology（防务技术） 2023年第3期21卷 176-183页

作者： Yang Liu Cong-qing Wang Yong-jun Zhou College of Automation Engineering Nanjing University of Aeronautics and AstronauticsNanjing211106China Science and Technology on Near-Surface Detection Laboratory Wuxi214000China

Automated detection of military people based on the images in different environments plays an important role in accurately completing military *** the equipment gradually moving towards intelligence,unmanned aerial vehicles(UAVs)will be widely used for integrated reconnaissance/attack in the *** lightweight and compact design of the small UAV allows it to travel through dense forests and other environments to capture images with its convenient ***,as the camouflage has been designed to blend in with surroundings,which greatly reduces the probability of the target being ***,the lack of training data on camouflaged people detection will inhibit the training of a deep *** address these problems,a novel semi-supervised camouflaged military people detection network is proposed to automatically detect the target from the *** this paper,the camouflaged object detection dataset(COD10K)is first supplemented according to our mission requirements,then the edge attention is utilized to enhance the boundaries based on search identification ***,a semi-supervised learning strategy is presented to take advantage of the unlabeled data which can alleviate insufficient data and improve the detection *** demonstrate that the proposed semi-supervised search identification network(Semi-SINet)performs well in camouflaged people detection compared with other object detection methods.

关键词： network environments utilized

来源：评论

学校读者我要写书评

暂无评论

A Hyperspectral Nonlinear Unmixing Network for nearshore Underwater Target detection 5

A Hyperspectral Nonlinear Unmixing Network for Nearshore Und...

引用

5th International Seminar on Artificial Intelligence, Networking and Information technology, AINIT 2024

作者： Liu, Jiaxuan Qi, Jiahao Zhong, Ping Jiang, Junhe National University of Defense Technology Key Laboratory of Science and Technology On Automatic Target Recognition Changsha China Science and Technology On Near-Surface Detection Laboratory Wuxi China

ISBN: (纸本)9798350385557

The existing hyperspectral underwater target detection algorithms are ineffective in adapting to nearshore environments, owing to the difficulty of extracting target signals from cluttered nearshore backgrounds combined with weak target signals. To this end, hyperspectral unmixing (HU) is well suited in this paper for nearshore underwater target detection (UTD) task to address these issues, by extracting background endmembers and separating out weak target signals simultaneously. Specifically, we propose a novel nonlinear unmixing network for nearshore underwater target detection (NUN-UTD). It comprises an unsymmetrical autoencoder and a target prior preservation module to adapt to the complex nearshore environment. Experiments demonstrate that NUN-UTD achieves the best detection performances and visual results compared with other state-of-the-art methods. © 2024 IEEE.

关键词： Learning systems

来源：评论

学校读者我要写书评

暂无评论

An Improved Channel Assignment and Topology Control Algorithm in Cognitive Radio Networks 3rd

An Improved Channel Assignment and Topology Control Algorith...

引用

3rd International Conference on Advanced Intelligent Technologies, ICAIT 2022

作者： Yuelin, Xu Kai, Ding Xingchun, Xu Science and Technology On Near-Surface Detection Laboratory Wuxi214035 China

ISBN: (纸本)9789819952021

Cognitive radio has emerged as a promising technology to exploit the unused portions of spectrum in an opportunistic manner. Spectrum assignment is a key mechanism that limits the interference between second users and primary users, enabling a more efficient usage of the wireless spectrum. The licensed channels are randomly occupied by primary users in the cognitive radio networks. The channel occupancy time of the primary user in different channels is different too. By assuming it is negative exponentially distributed and combining the channel selection criteria of the robust topology control algorithm (CRTCA), an improved channel assignment algorithm (ICAA) was proposed. The simulation results indicate that the channel assignment method of considering the channel occupancy time of primary user is more than CRTCA in throughput and less than CRTCA in collision rate. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Cognitive radio

来源：评论

学校读者我要写书评

暂无评论

Fractional Gradient Descent Algorithm for Nonlinear Additive Systems Using Weierstrass Approximation Method 3rd

Fractional Gradient Descent Algorithm for Nonlinear Additive...

引用

3rd International Conference on Advanced Intelligent Technologies, ICAIT 2022

作者： Rong, Yingjiao Peng, Fei Lv, Rongqi Li, Shanshan Science and Technology on Near-Surface Detection Laboratory Wuxi214028 China

ISBN: (纸本)9789819952021

In this study, a fractional gradient descent (FGD) algorithm for nonlinear additive systems with unknown structure is proposed. The nonlinear model is turned into a linear-parameter model using the Weierstrass approximation approach. Then, a FGD algorithm is created to estimate all of the system's unknown parameters. A preconditioned matrix approach is also presented to improve the FGD algorithm's convergence rate. To demonstrate the effectiveness of the proposed algorithms, simulation examples are provided. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Parameter estimation

来源：评论

学校读者我要写书评

暂无评论

Scenario Modeling-Aided AP Placement Optimization Method for Indoor Localization and Network Access

引用

China Communications 2024年第3期21卷 37-50页

作者： Pan Hao Chen Yu Qi Xiaogang Liu Meili School of Mathematics and Statistics Xidian UniversityXi’an 710126China Science and Technology on Near-Surface Detection Laboratory Wuxi 214035China School of Software Nanchang Hangkong UniversityNanchang 330063China

Owing to the ubiquity of wireless networks and the popularity of WiFi infrastructures,received signal strength(RSS)-based indoor localization systems have received much *** placement of access points(APs)significantly influences localization accuracy and network ***,the indoor scenario and network access are not fully considered in previous AP placement optimization *** study proposes a practical scenario modelingaided AP placement optimization method for improving localization accuracy and network *** order to reduce the gap between simulation-based and field measurement-based AP placement optimization methods,we introduce an indoor scenario modeling and Gaussian process-based RSS prediction *** that,the localization and network access metrics are implemented in the multiple objective particle swarm optimization(MOPSO)solution,Pareto front criterion and virtual repulsion force are applied to determine the optimal AP ***,field experiments demonstrate the effectiveness of the proposed indoor scenario modeling method and RSS prediction model.A thorough comparison confirms the localization and network access improvement attributed to the proposed anchor placement method.

关键词： indoor localization MOPSO network access RSS prediction

来源：评论

学校读者我要写书评

暂无评论

An adaptive formation path following flight control method for fixed-wing UAV based on dynamic target point 3

An adaptive formation path following flight control method f...

引用

2024 3rd International Conference on Aerospace and Control Engineering, ICoACE 2024

作者： Yang, Huitao Wen, Qiuqiu Peng, Fei School of Aerospace Engineering Beijing Institute of Technology Beijing100081 China Science and Technology on Near-Surface Detection Laboratory Jiangsu214035 China

Aiming at the shortcomings of the current fixed-wing UAV formation path tracking control method, which is complicated in design, poorly adapted to variable curvature paths, and low in accuracy, this paper proposes an adaptive formation path-following flight control method based on dynamic target points. First, based on the leader-follower architecture, this paper designs a generalized lateral control method for fixed-wing UAVs to realize the tracking of dynamic target points. Then, the leader UAV realizes independent path-following by selecting dynamic target points on the path. At the same time, the follower calculates the target points according to the formation parameters and the leader's state and additionally introduces a target distance control strategy to realize adaptive 2D formation keeping. Finally, mathematical simulations are carried out in this paper for verification. The simulation results show that this method can simultaneously take care of stable formation keeping and accurate path-following for fixed-wing UAV clusters. Compared with the existing methods, it can reduce the design difficulty, significantly improve the accuracy of complex path-following with variable curvature, and has good adaptive characteristics. © 2025 Institute of Physics Publishing. All rights reserved.

关键词： Flight paths

来源：评论

学校读者我要写书评

暂无评论

The posterior selection method for hyperparameters in regularized least squares method

引用

Control Theory and technology 2024年第2期22卷 184-194页

作者： Yanxin Zhang Jing Chen Yawen Mao Quanmin Zhu School of Science Jiangnan UniversityWuxi214122JiangsuChina The Science and Technology on Near-Surface Detection Laboratory Wuxi214122JiangsuChina Department of Engineering Design and Mathematics University of the West of EnglandBristolBS161QYUK

The selection of hyperparameters in regularized least squares plays an important role in large-scale system identification. The traditional methods for selecting hyperparameters are based on experience or marginal likelihood maximization method, which are inaccurate or computationally expensive. In this paper, two posterior methods are proposed to select hyperparameters based on different prior knowledge (constraints), which can obtain the optimal hyperparameters using the optimization theory. Moreover, we also give the theoretical optimal constraints, and verify its effectiveness. Numerical simulation shows that the hyperparameters and parameter vector estimate obtained by the proposed methods are the optimal ones.

关键词： Regularization method Hyperparameter System identification Least squares algorithm

来源：评论

学校读者我要写书评

暂无评论

RDSM: Underwater multi-AUV relay deployment and selection mechanism in 3D space

引用

Cognitive Robotics 2024年第1期4卷 204-216页

作者： Liu, Yafei Liu, Na Li, Hao Jiang, Yi zhu, Junwu CSSC Ocean Exploration Technology Institute Co. Ltd Jiangsu 214153 China College of Information Engineering Yangzhou University Jiangsu 225000 China Science and Technology on Near-Surface Detection Laboratory Jiangsu 214000 China

Underwater Wireless Sensor Networks (UWSNs) are widely used in naval military field and marine resource exploration. However, challenges such as resource inefficiency and unbalanced energy consumption severely hinder their practical applications. In this paper, we establish a model of underwater multi-hop wireless sensor network with multiple AUVs as relay nodes, which describes the data transmission process within the network. Based on this, an underwater multi-AUV Relay Deployment and Selection Mechanism in 3D space (RDSM) is proposed to achieve efficient underwater networking. Specifically, the RDSM includes the following key components. Firstly, an optimized relay node deployment strategy (RNDS) is used to deploy AUV nodes to effectively ensure network connectivity. Compared with traditional methods, this strategy has unique advantages in considering underwater space characteristics and can better adapt to the complex underwater environment. Secondly, a new utility function is constructed by integrating factors such as throughput, energy consumption, and load. The relay selection strategy based on utility maximization (RSS-UM) is used to select the next-hop relay node. This strategy is innovative in improving relay selection efficiency and optimizing network performance. Finally, in response to the problem of rapid energy consumption of relay nodes close to the base station, a power adjustment scheme is introduced to achieve a balance in node energy consumption, which is of great significance for prolonging network lifetime and improving overall stability. Experimental results show that compared with existing methods, the proposed mechanism achieves high utility and throughput, while maintaining balanced node energy consumption. © 2024

关键词： Sensor nodes

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：