检索结果-内蒙古大学图书馆

Delayed Memory Unit: Modeling Temporal Dependency Through Delay Gate

IEEE Transactions on Neural Networks and Learning Systems 2024年第6期36卷 10808-10818页

作者： Pengfei Sun Jibin Wu Malu Zhang Paul Devos Dick Botteldooren Department of Information Technology WAVES Research Group Ghent University Ghent Belgium Department of Data Science and Artificial Intelligence Department of Computing The Hong Kong Polytechnic University Hong Kong SAR China School of Computer Science and Engineering University of Electronic Science and Technology of China Chengdu China

Recurrent neural networks (RNNs) are widely recognized for their proficiency in modeling temporal dependencies, making them highly prevalent in sequential data processing applications. Nevertheless, vanilla RNNs are confronted with the well-known issue of gradient vanishing and exploding, posing a significant challenge for learning and establishing long-range dependencies. Additionally, gated RNNs tend to be over-parameterized, resulting in poor computational efficiency and network generalization. To address these challenges, this article proposes a novel delayed memory unit (DMU). The DMU incorporates a delay line structure along with delay gates into vanilla RNN, thereby enhancing temporal interaction and facilitating temporal credit assignment. Specifically, the DMU is designed to directly distribute the input information to the optimal time instant in the future, rather than aggregating and redistributing it over time through intricate network dynamics. Our proposed DMU demonstrates superior temporal modeling capabilities across a broad range of sequential modeling tasks, utilizing considerably fewer parameters than other state-of-the-art gated RNN models in applications such as speech recognition, radar gesture recognition, ECG waveform segmentation, and permuted sequential (PS) image classification.

关键词： Logic gates Delays Delay lines Recurrent neural networks Long short term memory Computational modeling Computational efficiency Transformers Sun Speech recognition

来源：评论

学校读者我要写书评

暂无评论

Leveraging Contaminated datasets to Learn Clean-data Distribution with Purified Generative Adversarial Networks

arXiv

引用

arXiv 2023年

作者： Tian, Bowen Su, Qinliang Yu, Jianxing School of Computer Science and Engineering Sun Yat-sen University Guangzhou China Guangdong Key Laboratory of Big Data Analysis and Processing Guangzhou China School of Artificial Intelligence Sun Yat-sen University Guangdong China

Generative adversarial networks (GANs) are known for their strong abilities on capturing the underlying distribution of training instances. Since the seminal work of GAN, many variants of GAN have been proposed. However, existing GANs are almost established on the assumption that the training dataset is clean. But in many real-world applications, this may not hold, that is, the training dataset may be contaminated by a proportion of undesired instances. When training on such datasets, existing GANs will learn a mixture distribution of desired and contaminated instances, rather than the desired distribution of desired data only (target distribution). To learn the target distribution from contaminated datasets, two purified generative adversarial networks (PuriGAN) are developed, in which the discriminators are augmented with the capability to distinguish between target and contaminated instances by leveraging an extra dataset solely composed of contamination instances. We prove that under some mild conditions, the proposed PuriGANs are guaranteed to converge to the distribution of desired instances. Experimental results on several datasets demonstrate that the proposed PuriGANs are able to generate much better images from the desired distribution than comparable baselines when trained on contaminated datasets. In addition, we also demonstrate the usefulness of PuriGAN on downstream applications by applying it to the tasks of semi-supervised anomaly detection on contaminated datasets and PU-learning. Experimental results show that PuriGAN is able to deliver the best performance over comparable baselines on both tasks1 Copyright © 2023, The Authors. All rights reserved.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

GCD-Sampling: A General Cross-scale Decoupled Sampling for Point Cloud 39

GCD-Sampling: A General Cross-scale Decoupled Sampling for P...

引用

39th Annual AAAI Conference on Artificial Intelligence, AAAI 2025

作者： Dai, Tao Wang, Yanzi Xiong, Jianyu Zha, Yaohua Xia, Shu-Tao Zhu, Zexuan College of Computer Science and Software Engineering Shenzhen University China National Engineering Laboratory for Big Data System Computing Technology Shenzhen University China Tsinghua Shenzhen International Graduate School Tsinghua University China Shenzhen City Key Laboratory of Embedded System Design Shenzhen China

ISBN: (纸本)157735897X

Sampling strategy (e.g., fixed farthest point sampling) of point cloud has been an essential step for developing practical solutions in 3D computer vision tasks. Previous fixed sampling is simple, but suffer from suboptimal performance for downstream tasks. To adapt to target networks properly, adaptive sampling methods with trainable parameters have been recently developed to enhance the performance. However, existing adaptive sampling methods still suffer from the over-coupling problem of target network, and thus become model-specific, which limits their practical applications. To address this issue, we propose a novel general cross-scale decoupled sampling method (GCD-sampling) for point cloud, which consists of original feature cache, cross-scale feature fusion and convex combination learning for better feature extraction. To reduce the coupling relationship with the target task network, our method only utilizes the point cloud coordinates as the input and output of itself. Besides, we introduce an arbitrary scale structure to enable parameter sharing across multi-scale sampling in point cloud networks. Extensive experiments on different architectures demonstrate the effectiveness of our method over other existing adaptive sampling methods. © 2025, Association for the Advancement of Artificial Intelligence (***). All rights reserved.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

EPL: Empirical Prototype Learning for Deep Face Recognition

arXiv

引用

arXiv 2024年

作者： Fan, Weijia Wen, Jiajun Jia, Xi Shen, Linlin Zhou, Jiancan Li, Qiufu National Engineering Laboratory for Big Data System Computing Technology Shenzhen University Shenzhen518060 China Computer Vision Institute Shenzhen University Shenzhen518060 China School of Computer Science University of Birmingham Birmingham United Kingdom Aqara Lumi United Technology Coopration Shenzhen China

Prototype learning is widely used in face recognition, which takes the row vectors of coefficient matrix in the last linear layer of the feature extraction model as the prototypes for each class. When the prototypes are updated using the facial sample feature gradients in the model training, they are prone to being pulled away from the class center by the hard samples, resulting in decreased overall model performance. In this paper, we explicitly define prototypes as the expectations of sample features in each class and design the empirical prototypes using the existing samples in the dataset. We then devise a strategy to adaptively update these empirical prototypes during the model training based on the similarity between the sample features and the empirical prototypes. Furthermore, we propose an empirical prototype learning (EPL) method, which utilizes an adaptive margin parameter with respect to sample features. EPL assigns larger margins to the normal samples and smaller margins to the hard samples, allowing the learned empirical prototypes to better reflect the class center dominated by the normal samples and finally pull the hard samples towards the empirical prototypes through the learning. The extensive experiments on MFR, IJB-C, LFW, CFP-FP, AgeDB, and MegaFace demonstrate the effectiveness of EPL. Our code is available at https://***/WakingHours-GitHub/EPL. © 2024, CC BY.

关键词： Face recognition

来源：评论

学校读者我要写书评

暂无评论

SpatioTemporal Learning for Human Pose Estimation in Sparsely-Labeled Videos 39

SpatioTemporal Learning for Human Pose Estimation in Sparsel...

引用

39th Annual AAAI Conference on Artificial Intelligence, AAAI 2025

作者： Jiao, Yingying Wang, Zhigang Wu, Sifan Fan, Shaojing Liu, Zhenguang Xu, Zhuoyue Wu, Zheqi College of Computer Science and Technology Jilin University China Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Jilin University China College of Computer Science and Technology Zhejiang Gongshang University China School of Computing National University of Singapore Singapore The State Key Laboratory of Blockchain and Data Security Zhejiang University China Hangzhou High-Tech Zone (Binjiang) Institute of Blockchain and Data Security China

ISBN: (纸本)157735897X

Human pose estimation in videos remains a challenge, largely due to the reliance on extensive manual annotation of large datasets, which is expensive and labor-intensive. Furthermore, existing approaches often struggle to capture long-range temporal dependencies and overlook the complementary relationship between temporal pose heatmaps and visual features. To address these limitations, we introduce STDPose, a novel framework that enhances human pose estimation by learning spatiotemporal dynamics in sparsely-labeled videos. STDPose incorporates two key innovations: 1) A novel Dynamic-Aware Mask to capture long-range motion context, allowing for a nuanced understanding of pose changes. 2) A system for encoding and aggregating spatiotemporal representations and motion dynamics to effectively model spatiotemporal relationships, improving the accuracy and robustness of pose estimation. STDPose establishes a new performance benchmark for both video pose propagation (i.e., propagating pose annotations from labeled frames to unlabeled frames) and pose estimation tasks, across three large-scale evaluation datasets. Additionally, utilizing pseudo-labels generated by pose propagation, STDPose achieves competitive performance with only 26.7% labeled data. Copyright © 2025, Association for the Advancement of Artificial Intelligence (***). All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Oriented Object Detection in Aerial Images with Double Midpoint Vectors

Oriented Object Detection in Aerial Images with Double Midpo...

引用

International Conference on computer and Communications (ICCC)

作者： Xiaoke Zhu Qi Hu Xiaopan Chen Caihong Yuan Minghao Li School of Computer and Information Engineering Henan University Kaifeng China Henan Engineering Research Center of Intelligent Technology and Application Henan University Henan Key Laboratory of Big Data Analysis and Processing Henan University Kaifeng China

In recent years, oriented object detection technology has made significant progress. Most oriented object detectors represent oriented bounding boxes by adding additional angle information to the horizontal bounding box representation, which leads to serious angle periodicity issues. To address these issues, this paper proposes double midpoint vectors (DMvectors) to represent oriented objects. For any oriented bounding box, we align the center point with the origin in the Cartesian coordinate system, and evenly distribute the midpoints of the box edges across the four quadrants. We use the midpoint of the first and second quadrants, as well as the width and height of the box, to represent the oriented bounding box. DMvectors solve the periodic problem of angles by implicitly including angle information in vectors. To demonstrate the effectiveness of our proposed method, we conducted extensive experiments on two common remote sensing datasets (DOTA and HRSC2016), and our method achieved competitive results.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Finding Important Edges in Networks through Local Information 7

Finding Important Edges in Networks through Local Informatio...

引用

7th International Conference on computer and Communications, ICCC 2021

作者： Yu, En-Yu Fu, Yan Zhou, Jun-Lin Chen, Duan-Bing School of Computer Science and Engineering University of Electronic Science and Technology of China Chengdu611731 China Union Big Data Tech. Inc. Chengdu610041 China

ISBN: (纸本)9781665409506

In transportation, communication, social and other real complex networks, some critical edges act a pivotal part in controlling the flow of information and maintaining the integrity of the structure. Due to the importance of critical edges in theoretical studies and practical applications, the identification of critical edges gradually become a hot topic in current researches. Considering the overlap of communities in the neighborhood of edges, a novel and effective metric named subgraph overlap (SO) is proposed to quantifying the significance of edges. The experimental results show that SO outperforms all benchmarks in identifying critical edges which are crucial in maintaining the integrity of the structure and functions of networks. © 2021 IEEE.

关键词： Complex networks

来源：评论

学校读者我要写书评

暂无评论

A Low-Code Development Framework for Constructing Industrial Apps 15th

A Low-Code Development Framework for Constructing Industrial...

引用

15th CCF Conference on computer Supported Cooperative Work and Social Computing, Chinese CSCW 2020

作者： Wang, Jingyue Qi, Binhang Zhang, Wentao Sun, Hailong SKLSDE Lab School of Computer Science and Engineering Beihang University Beijing100191 China Beijing Advanced Innovation Center for Big Data and Brain Computing Beijing100191 China

ISBN: (纸本)9789811625398

With the advent of the Industry 4.0, intelligent manufacturing has become a technological highland to conquer in the process of enterprise digitalization. As the core competitiveness of intelligent manufacturing, industrial apps, with new features such as customization and lightweight, has emerged as a new type of industrial software. Traditional development methods and tools can hardly meet the large demand of industrial software on account of its long development cycles while low-code development can greatly improve the productivity of industrial software, lower the barriers and reduce costs for development. Therefore, the research and application of low-code development for industrial apps has received much attention. Industrial Internet platforms such as Siemens, OutSystems have successively launched low-code tools. However, there is still a lack of an open, unified low-code development framework in industry. In response to the above problems, we propose a low-code framework to develop industrial apps quickly and easily, which paves the way for leveraging the crowd intelligence of worldwide developers to improve the productivity of developing industrial apps. Based on BPMN2.0 and Apache Activiti engine, this framework provides drag-and-drop process design, one-click process deployment and operation, data monitoring and other functions. In this paper, we present a prototype system of a low-code development framework and demonstrate its functions through a use case of developing a predictive maintenance application. Finally, the aircraft turbine life data is used to verify the effectiveness of the system. © 2021, Springer Nature Singapore Pte Ltd.

关键词： Productivity

来源：评论

学校读者我要写书评

暂无评论

Near-Field Integrated Sensing and Communication with Extremely Large-Scale Antenna Array

arXiv

引用

arXiv 2024年

作者： Hua, Haocheng Xu, Jie Zhang, Rui The Chinese University of Hong Kong Guandong Shenzhen518172 China The School of Science and Engineering Shenzhen Research Institute of Big Data The Chinese University of Hong Kong Guangdong Shenzhen518172 China The Department of Electrical and Computer Engineering National University of Singapore Singapore117583 Singapore

This paper studies a near-field integrated sensing and communication (ISAC) system with extremely large-scale antenna array (ELAA), in which a base station (BS) deployed with a very large number of antennas transmits wireless signals to communicate with multiple communication users (CUs) and simultaneously uses the echo signals to localize multiple point targets in the three-dimension (3D) space. To balance the performance tradeoff between near-field communication and 3D target localization, we design the transmit covariance matrix at the BS to optimize the localization performance while ensuring the signal-to-interference-plus-noise ratio (SINR) constraints at individual CUs. In particular, we formulate three design problems by considering different 3D localization performance metrics, including minimizing the sum Cramér-Rao bound (CRB) for estimating 3D locations, maximizing the minimum target illumination power, and maximizing the minimum target echo signal power. Although the three design problems are non-convex in general, we obtain their global optimal solutions via the technique of semi-definite relaxation (SDR) by proving the tightness of such relaxations. It is rigorously shown that the optimal solutions to the three problems have low-rank structures depending on the sensing and communication channel matrices, which can be exploited to greatly reduce the computational complexity of the SDR-based solutions. Interestingly, we find that in the special case with a single collocated target/CU present towards the middle of a symmetric uniform planar array (UPA), the optimal solutions to the three problems become identical to the SINR-maximization design and have a closed form, while in other cases they can be different in general. Besides, when the target/CU moves away from the transmitter/receiver, the CRB may first decrease and then increase. These two phenomena differ from those in the far-field scenario. Numerical results show the benefits of the proposed des

关键词： Beamforming

来源：评论

学校读者我要写书评

暂无评论

FLeS: A Federated Learning-Enhanced Semantic Communication Framework for Mobile AIGC-Driven Human Digital Twins

引用

IEEE Network 2025年

作者： Okegbile, Samuel D. Gao, Haoran Talabi, Oluwasegun Cai, Jun Yi, Changyan Niyato, Dusit Shen, Xuemin University of the Fraser Valley School of Computing AbbotsfordBCV2S 7M8 Canada Concordia University Department of Electrical and Computer Engineering MontrealQCH3G 1M8 Canada Nanjing University of Aeronautics and Astronautics College of Computer Science and Technology Jiangsu Nanjing211106 China Nanyang Technological University College of Computing and Data Science Singapore University of Waterloo Department of Electrical and Computer Engineering Canada

Innovative mobile artificial intelligence-generated content (AIGC) can support the evolution and updating processes of virtual twins (VTs) in human digital twin (HDT) systems. With a reliable and efficient automatic data generation process, the requirement for a timely physical-to-virtual synchronization in HDT can be satisfied. While such an AIGC-enabled HDT system can facilitate modelling high fidelity VTs, generating content that represents the true states in the physical environment and providing timely customized services, it may suffer from a poor understanding of contexts, a lack of creativity, and various security and privacy concerns. In this paper, we propose a novel framework, which integrates federated learning (FL) and semantic communication (SemCom) to enhance performance in the AIGC-enabled HDT system while improving accuracy and convergence properties. First, we present a holistic architectural framework for the proposed FL-enhanced SemCom (FLeS) solution for mobile AIGC-enabled HDT systems and discuss its design requirements and challenges. We later present key technologies necessary to realize the FLeS solution, followed by elaborating on important technical issues to suggest future directions. Experimental results demonstrate that FLeS not only facilitates reliable and personalized content generation but also shows better performance when compared to existing solutions. © 1986-2012 IEEE.

关键词： Federated learning

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：