检索结果-内蒙古大学图书馆

Music source separation via hybrid waveform and spectrogram based generative adversarial network

Multimedia Tools and Applications 2024年 1-15页

作者： Wu, Qiuxia Deng, Haipeng Hu, Kun Wang, Zhiyong School of Software Engineering South China University of Technology Guangdong Province Guangzhou China School of Computer Science The University of Sydney SydneyNSW Australia

Music source separation aims to disentangle individual sources from the mixture of musical signals. Existing generative adversarial network (GAN) based methods generally work on the spectrogram domain only. However, this practice ignores the patterns from the waveform domain, which are more informative for modelling some categories of sources. In this paper, we propose a fully hybrid GAN framework to integrate knowledge from both domains. In particular, the generator formulates acoustical patterns from waveform and spectrogram domains, while the discriminator provides discriminative information based on the local patch-level spectrograms such that the generator can produce more plausible separation results. Furthermore, to enhance the quality of estimated sources, we devise a perceptual spectrogram loss term, which is a complement of the waveform-level loss. The proposed method is evaluated on two widely used music source separation datasets, producing music sources of high signal-to-distortion ratio (12.03 in MIR-1K dataset and 8.08 in MUSDB18 dataset). These results demonstrate the superiority of the proposed method compared with the state-of-the-art methods. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

MCARNet: Mixed Convolution-Attention-Residual Network for Hyperspectral Image Classification of Cholangiocarcinoma 12

MCARNet: Mixed Convolution-Attention-Residual Network for Hy...

引用

12th International Conference on Computing and Pattern Recognition, ICCPR 2023

作者： Xue, Jiaqi He, Wangquan Ye, Songxin Liang, Hong Jia, Sen College of Computer Science and Software Engineering Shenzhen University China

ISBN: (纸本)9798400707988

Hyperspectral is a well-established form of data in the medical field and plays an irreplaceable role, reflecting not only the size, shape and structure of the target pathological areas, but also the differences in spectral between the chemical composition of the targets. Therefore, hyperspectral image(HSI) can provide more abundant information for medical practitioners when performing pathology analysis. In recent years, while two powerful representational learning methods, which have been fully developed in the long term in deep learning, namely convolution and attention mechanism, have both made significant contributions to the pathological analysis of hyperspectral medical images. However, they have also revealed respective shortcomings in previous works, such as insufficient support of the biological basis of convolution and its weak ability to capture global information, the insufficient ability of attention mechanism to acquire spatial texture features, and both have over-reliance on sample quantity as well as quality. Therefore, this article proposes a Mixed Convolution-Attention-Residual(MCAR) model and a corresponding MCARNet network for the HSI classification task of cholangiocarcinoma by effectively combining multi-branch convolution, attention mechanism and residual structure. We also effectively validate that the proposed model can perform better in the task by parallel extraction of global with local features as well as the construction of residual structure under the condition of few training samples. © 2023 ACM.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Dependable Service-Oriented Design of Healthcare IoT 7

Dependable Service-Oriented Design of Healthcare IoT

引用

7th IEEE International Conference on Smart Internet of Things, SmartIoT 2023

作者： Alsaig, Alaa Alagar, Vangalur Concordia University Computer Science and Software Engineering Montreal Canada

ISBN: (纸本)9798350316575

Healthcare Internet of Things (HIoT) is transforming healthcare industry by providing large scale connectivity for medical devices, patients, physicians, clinical staff, and nurses who may use them remotely to facilitate real-time monitoring based on the information gathered from the connected medical devices. This network provides both opportunity and challenges for collecting relevant healthcare information and sharing it for innovative research and efficient healthcare provision. Patient-centric information such as their health status and medical devices used by them must be protected to provide safety and offer privacy, while healthcare knowledge should be shared in confidence by experts for healthcare innovation and timely treatment of patients. In this paper a dependable service-oriented 3-tier architecture of HIoT is proposed. Context-aware role-based access control scheme is discussed to ensure services in HIoT are dependable, in the sense that every service provided to any patient in the network, anywhere and at anytime is safe, secure, reliable, and the privacy of patient receiving it is protected. © 2023 IEEE.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

Economic Analysis of Joint Mobile Edge Caching and Peer Content Sharing 42

Economic Analysis of Joint Mobile Edge Caching and Peer Cont...

引用

42nd IEEE International Conference on computer Communications, INFOCOM 2023

作者： Jiang, Changkun Shenzhen University College of Computer Science and Software Engineering China

ISBN: (纸本)9798350334142

Mobile edge caching (MEC) allows edge devices to cache popular contents and deliver them to end-users directly, which can effectively alleviate increasing backbone loads and improve end-users' quality of service. Peer content sharing (PCS) enables edge devices to share the cached contents with others, and can further increase the caching efficiency. While many efforts have been made to content caching or sharing technologies, it remains open to study the complicated technical and economic interplay between both technologies. In this paper, we propose a joint MEC-PCS framework, and focus on capturing the strategic interactions between different types of edge devices. Specifically, we model their interactions as a non-cooperative game, where each device (player) can choose to be an agent (who caches and shares contents with others) or a requester (who doesn't cache but requests contents from others) of each content. We analyze the existence and uniqueness of the game equilibrium systematically under the generic usage-based pricing (for PCS). To address the incomplete information issue, we further design a behavior rule that allows players to achieve the equilibrium via a dynamic learning algorithm. Simulations show that the joint MEC-PCS framework can reduce the total system cost by 60%, compared with the benchmark pure caching system without PCS. © 2023 IEEE.

关键词： Economic analysis

来源：评论

学校读者我要写书评

暂无评论

Topology Discovery in Autonomic Networks 36

Topology Discovery in Autonomic Networks

引用

36th IEEE/IFIP Network Operations and Management Symposium, NOMS 2023

作者： Ghaderi, Parsa Atwood, J. William Narayanan, Lata Concordia University Computer Science and Software Engineering Montreal Canada

ISBN: (纸本)9781665477161

Access to a topological map is important to network management. Autonomic Networks have been developed, in part, because they provide a convenient, secure platform for the development of operations, administration and management (OAM) applications, and because they make it possible to substantially reduce the need for human input in many network management situations. Using the definition of Autonomic Networks that has been developed by the Network Management Research Group of the IRTF, and standardized by the ANIMA Working Group of the IETF, we present two methods for discovering and maintaining topological information within an Autonomic Network domain. Our first method is highly distributed, and is based on clustering, while the second method is more centralized, and takes advantage of the already-defined procedure for securely initializing the Autonomic Network. We implemented both solutions on a testbed configured with three different topologies. We report the relative performance of the proposed methods. © 2023 IEEE.

关键词： Network management

来源：评论

学校读者我要写书评

暂无评论

Intelligent obstacle avoidance algorithm for safe urban monitoring with autonomous mobile drones

引用

Journal of Electronic science and Technology 2024年第4期22卷 19-36页

作者： Didar Yedilkhan Abzal E.Kyzyrkanov Zarina A.Kutpanova Shadi Aljawarneh Sabyrzhan K.Atanov Department of Computer Engineering Astana IT UniversityAstana010000Kazakhstan Department of Computer and Software Engineering L.N.Gumilyov Eurasian National UniversityAstana010000Kazakhstan Department of Automation and Control Systems L.N.Gumilyov Eurasian National UniversityAstana010000Kazakhstan Department of Software Engineering Jordan University of Science and TechnologyIrbid22110Jordan

The growing field of urban monitoring has increasingly recognized the potential of utilizing autonomous technologies,particularly in drone *** deployment of intelligent drone swarms offers promising solutions for enhancing the efficiency and scope of urban condition *** this context,this paper introduces an innovative algorithm designed to navigate a swarm of drones through urban landscapes for monitoring *** primary challenge addressed by the algorithm is coordinating drone movements from one location to another while circumventing obstacles,such as *** algorithm incorporates three key components to optimize the obstacle detection,navigation,and energy efficiency within a drone ***,the algorithm utilizes a method to calculate the position of a virtual leader,acting as a navigational beacon to influence the overall direction of the ***,the algorithm identifies observers within the swarm based on the current *** further refine obstacle avoidance,the third component involves the calculation of angular velocity using fuzzy *** approach considers the proximity of detected obstacles through operational rangefinders and the target’s location,allowing for a nuanced and adaptable computation of angular *** integration of fuzzy logic enables the drone swarm to adapt to diverse urban conditions dynamically,ensuring practical obstacle *** proposed algorithm demonstrates enhanced performance in the obstacle detection and navigation accuracy through comprehensive *** results suggest that the intelligent obstacle avoidance algorithm holds promise for the safe and efficient deployment of autonomous mobile drones in urban monitoring applications.

关键词： Drone swarms Fuzzy logic Intelligent solution Smart city Urban monitoring

来源：评论

学校读者我要写书评

暂无评论

Vehicle logo recognition based on depth residual shrinkage network 3

Vehicle logo recognition based on depth residual shrinkage n...

引用

3rd International Conference on Artificial Intelligence and computer engineering, ICAICE 2022

作者： Lin, Zhaobo Computer Science and Software Engineering Fuzhou University 350000 China

ISBN: (纸本)9781510663473

So as to reduce the interference of noisy information and enhance the accuracy of car marker image recognition in real life, a car marker image recognition model based on deep residual shrinkage network is proposed. The model combines attention mechanisms and soft thresholding function on the basis of deep residual network, which is used to eliminate noise and redundant information in the data, thus reducing the interference of noise information and improving the accuracy of image recognition. The experiments are conducted by adding Gaussian noise and pretzel noise to the car logo images shared by HFUT-VL on Github to simulate the car logo images captured under realistic conditions, forming a dataset of 200 images for each car logo with a total of 16000 images, and then training the deep residual shrinkage network with deep residual network and SENet algorithm model on the data. Then, the deep residual shrinkage network is compared with the deep residual network and SENet algorithm model to train the data, and some of the car mark images are tested. The results of the experiment show that the method has better performance than other deep neural network methods. © 2023 SPIE.

关键词： Shrinkage

来源：评论

学校读者我要写书评

暂无评论

Multi-Agent Graph Reinforcement Learning based On-Demand Wireless Energy Transfer in Multi-UAV-aided IoT Network 21

Multi-Agent Graph Reinforcement Learning based On-Demand Wir...

引用

21st International Symposium on Modeling and Optimization in Mobile, Ad Hoc, and Wireless Networks, WiOpt 2023

作者： Zhao, Ze Yu Che, Yueling Luo, Sheng Wu, Kaishun Leung, Victor C.M. College of Computer Science and Software Engineering Shenzhen University China

ISBN: (纸本)9783903176553

This paper proposes a new on-demand wireless energy transfer (WET) scheme of multiple unmanned aerial vehicles (UAVs). Unlike the existing studies that simply pursuing the total or the minimum harvested energy maximization at the Internet of Things (IoT) devices, where the IoT devices’ own energy requirements are barely considered, we propose a new metric called the hungry-level of energy (HoE), which reflects the time-varying energy demand of each IoT device based on the energy gap between its required energy and the harvested energy from the UAVs. With the purpose to minimize the overall HoE of the IoT devices whose energy requirements are not satisfied, we optimally determine all the UAVs’ trajectories and WET decisions over time, under the practical mobility and energy constraints of the UAVs. Although the proposed problem is of high complexity to solve, by excavating the UAVs’ self-attentions for their collaborative WET, we propose the multi-agent graph reinforcement learning (MAGRL) based approach. Through the offline training of the MAGRL model, where the global training at the central controller guides the local training at each UAV agent, each UAV then distributively determines its trajectory and WET based on the well-trained local neural networks. Simulation results show that the proposed MAGRL-based approach outperforms various benchmarks for meeting the IoT devices’ energy requirements. © 2023 IFIP.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

ER-Net:Efficient Recalibration Network for Multi-ViewMulti-Person 3D Pose Estimation

引用

computer Modeling in engineering & sciences 2023年第8期136卷 2093-2109页

作者： Mi Zhou Rui Liu Pengfei Yi Dongsheng Zhou National and Local Joint Engineering Laboratory of Computer Aided Design School of Software EngineeringDalian UniversityDalian116622China School of Computer Science and Technology Dalian University of TechnologyDalian116024China

Multi-view multi-person 3D human pose estimation is a hot topic in the field of human pose estimation due to its wide range of application *** the introduction of end-to-end direct regression methods,the field has entered a new stage of ***,the regression results of joints that are more heavily influenced by external factors are not accurate enough even for the optimal *** this paper,we propose an effective feature recalibration module based on the channel attention mechanism and a relative optimal calibration strategy,which is applied to themulti-viewmulti-person 3D human pose estimation task to achieve improved detection accuracy for joints that are more severely affected by external ***,it achieves relative optimal weight adjustment of joint feature information through the recalibration module and strategy,which enables the model to learn the dependencies between joints and the dependencies between people and their corresponding *** call this method as the Efficient Recalibration Network(ER-Net).Finally,experiments were conducted on two benchmark datasets for this task,Campus and Shelf,in which the PCP reached 97.3% and 98.3%,respectively.

关键词： Multi-view multi-person pose estimation attention mechanism computer vision

来源：评论

学校读者我要写书评

暂无评论

Boosting Write Performance of KV Stores: An NVM -Enabled Storage Collaboration Approach 40

Boosting Write Performance of KV Stores: An NVM -Enabled Sto...

引用

40th IEEE International Conference on Data engineering, ICDE 2024

作者： Wang, Yi He, Jiajian Sun, Kaoyi Dong, Yunhao Chen, Jiaxian Ma, Chenlin Zhou, Amelie Chi Mao, Rui College of Computer Science and Software Engineering Shenzhen University China Hong Kong Baptist University Department of Computer Science Hong Kong

ISBN: (纸本)9798350317152

As the most common data structure for key-value stores, LogStructured Merge Tree (LSM-tree) can eliminate random write operations and keep acceptable read performance. However, write stall and write amplification introduced by the leveled compaction of LSM-tree significantly degrade the system performance. The emerging non-volatile memory (NVM) provides byte-addressable access and low-latency data persistence. Integrating DIMM-interface NVM in the design of the LSM-tree can potentially alleviate the write stall and write amplification issue, as the access speed of NVM is several orders of magnitude faster than hard disk drives or flash memory-based solid-state drives. This hybrid storage should be carefully designed, requiring new architectural and key-value structural support. This paper presents ZigZagDB, an NVM-enabled data man-agement scheme for LSM-tree-based key-value stores. ZigZagDB adds additional layers of key-value stores and uses non-volatile memory as the storage media to hold these additional layers of data. The newly designed key-value stores alternately access the data from either SSD or NVM. This 'ZigZag' shape of storage collaboration and synchronization can benefit write efficiency and space utilization. By utilizing the NVM with very limited capacity, the redesigned organization of LSM-tree can effectively solve the write stall and write amplification issue. We demonstrate the viability of the proposed ZigZagDB using a set of extensive experiments. Experimental results show that ZigZagDB can significantly reduce the write amplification and boost the throughput in comparison with representative schemes. © 2024 IEEE.

关键词： Flash memory

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：