检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Ding, Chuntao Cao, Xu Xie, Jianhang Fan, Linlin Wang, Shangguang Lu, Zhichao School of Artificial Intelligence Beijing Normal University Beijing China Key Laboratory of Big Data & Artificial Intelligence in Transportation Ministry of Education with School of Computer Science and Technology Beijing Jiaotong University Beijing100044 China School of Computer and Information Engineering Henan Normal University Henan453007 China State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing China Department of Computer Science City University of Hong Kong HKSAR Hong Kong

Efficient fine-tuning of pre-trained convolutional neural network (CNN) models using local data is essential for providing high-quality services to users using ubiquitous and resource-limited Internet of Things (IoT) devices. Low-Rank Adaptation (LoRA) fine-tuning has attracted widespread attention from industry and academia because it is simple, efficient, and does not incur any additional reasoning burden. However, most of the existing advanced methods use LoRA to fine-tune Transformer, and there are few studies on using LoRA to fine-tune CNN. The CNN model is widely deployed on IoT devices for application due to its advantages in comprehensive resource occupancy and performance. Moreover, IoT devices are widely deployed outdoors and usually process data affected by the environment (such as fog, snow, rain, etc.). The goal of this paper is to use LoRA technology to efficiently improve the robustness of the CNN model. To this end, this paper first proposes a strong, robust CNN fine-tuning method for IoT devices, LoRA-C, which performs low-rank decomposition in convolutional layers rather than kernel units to reduce the number of fine-tuning parameters. Then, this paper analyzes two different rank settings in detail and observes that the best performance is usually achieved when α/r is a constant in either standard data or corrupted data. This discovery provides experience for the widespread application of LoRA-C. Finally, this paper conducts many experiments based on pre-trained models. Experimental results on CIFAR-10, CIFAR-100, CIFAR-10-C, and Icons50 datasets show that the proposed LoRA-Cs outperforms standard ResNets. Specifically, on the CIFAR-10-C dataset, the accuracy of LoRA-C-ResNet-101 achieves 83.44% accuracy, surpassing the standard ResNet-101 result by +9.5%. On the Icons-50 dataset, the accuracy of LoRA-C-ResNet-34 achieves 96.9% accuracy, surpassing the standard ResNet-34 result by +8.48%. In addition, compared with full parameter fine-tuning, LoRA-

关键词： Quality of service

来源：评论

学校读者我要写书评

暂无评论

Augmenting Prototype Network with TransMix for Few-shot Hyperspectral Image Classification

arXiv

引用

arXiv 2024年

作者： Liu, Chun Yang, Longwei Dong, Dongmei Li, Zheng Yang, Wei Han, Zhigang Wang, Jiayao School of Computer and Information Engineering Henan Key Laboratory of Big Data Analysis and Processing Henan Engineering Laboratory of Spatial Information Processing Henan Industrial Technology Academy of Spatio-Temporal Big Data Henan University Zhengzhou450046 China Beijing Electro-Mechanical Engineering Institute Beijing100074 China College of Geography and Environmental Science Henan Industrial Technology Academy of Spatio-Temporal Big Data Henan University Zhengzhou450046 China

Few-shot hyperspectral image classification aims to identify the classes of each pixel in the images by only marking few of these pixels. And in order to obtain the spatial-spectral joint features of each pixel, the fixed-size patches centering around each pixel are often used for classification. However, observing the classification results of existing methods, we found that boundary patches corresponding to the pixels which are located at the boundary of the objects in the hyperspectral images, are hard to classify. These boundary patchs are mixed with multi-class spectral information. Inspired by this, we propose to augment the prototype network with TransMix for few-shot hyperspectrial image classification(APNT). While taking the prototype network as the backbone, it adopts the transformer as feature extractor to learn the pixel-to-pixel relation and pay different attentions to different pixels. At the same time, instead of directly using the patches which are cut from the hyperspectral images for training, it randomly mixs up two patches to imitate the boundary patches and uses the synthetic patches to train the model, with the aim to enlarge the number of hard training samples and enhance their diversity. And by following the data agumentation technique TransMix, the attention returned by the transformer is also used to mix up the labels of two patches to generate better labels for synthetic patches. Compared with existing methods, the proposed method has demonstrated sate of the art performance and better robustness for few-shot hyperspectral image classification in our experiments. All the codes are available at https://***/HENULWY/APNT. Copyright © 2024, The Authors. All rights reserved.

关键词： Pixels

来源：评论

学校读者我要写书评

暂无评论

Efficient and Robust Continual Graph Learning for Graph Classification in Biology

arXiv

引用

arXiv 2024年

作者： Zhang, Ding Downer, Jane Chen, Can Wang, Ren The Department of Computer Science Northwestern University EvanstonIL60208 United States The Department of Computer Science Illinois Institute of Technology ChicagoIL60616 United States The School of Data Science and Society The Department of Mathematics University of North Carolina at Chapel Hill Chapel HillNC27599 United States The Department of Electrical and Computer Engineering Illinois Institute of Technology ChicagoIL60616 United States

Graph classification is essential for understanding complex biological systems, where molecular structures and interactions are naturally represented as graphs. Traditional graph neural networks (GNNs) perform well on static tasks but struggle in dynamic settings due to catastrophic forgetting. We present Perturbed and Sparsified Continual Graph Learning (PSCGL), a robust and efficient continual graph learning framework for graph data classification, specifically targeting biological datasets. We introduce a perturbed sampling strategy to identify critical data points that contribute to model learning and a motif-based graph sparsification technique to reduce storage needs while maintaining performance. Additionally, our PSCGL framework inherently defends against graph backdoor attacks, which is crucial for applications in sensitive biological contexts. Extensive experiments on biological datasets demonstrate that PSCGL not only retains knowledge across tasks but also enhances the efficiency and robustness of graph classification models in biology. Copyright © 2024, The Authors. All rights reserved.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

EfficientFace: An Efficient Deep Network with Feature Enhancement for Accurate Face Detection

arXiv

引用

arXiv 2023年

作者： Wang, Guangtao Li, Jun Wu, Zhijian Xu, Jianhua Shen, Jifeng Yang, Wankou School of Computer and Electronic Information Nanjing Normal University Nanjing210023 China School of Data Science and Engineering East China Normal University Shanghai200062 China School of Electrical and Information Engineering Jiangsu University Zhenjiang212013 China School of Automation Southeast University Nanjing210096 China

In recent years, deep convolutional neural networks (CNN) have significantly advanced face detection. In particular, lightweight CNN-based architectures have achieved great success due to their low-complexity structure facilitating real-time detection tasks. However, current lightweight CNN-based face detectors trading accuracy for efficiency have inadequate capability in handling insufficient feature representation, faces with unbalanced aspect ratios and occlusion. Consequently, they exhibit deteriorated performance far lagging behind the deep heavy detectors. To achieve efficient face detection without sacrificing accuracy, we design an efficient deep face detector termed EfficientFace in this study, which contains three modules for feature enhancement. To begin with, we design a novel cross-scale feature fusion strategy to facilitate bottom-up information propagation, such that fusing low-level and high-level features is further strengthened. Besides, this is conducive to estimating the locations of faces and enhancing the descriptive power of face features. Secondly, we introduce a Receptive Field Enhancement module to consider faces with various aspect ratios. Thirdly, we add an Attention Mechanism module for improving the representational capability of occluded faces. We have evaluated EfficientFace on four public benchmarks and experimental results demonstrate the appealing performance of our method. In particular, our model respectively achieves 95.1% (Easy), 94.0% (Medium) and 90.1% (Hard) on validation set of WIDER Face dataset, which is competitive with heavyweight models with only 1/15 computational costs of the state-of-the-art MogFace detector. Copyright © 2023, The Authors. All rights reserved.

关键词： Aspect ratio

来源：评论

学校读者我要写书评

暂无评论

Cyber Twin Technology for AI-Driven Real-Time Software Security in IoT Ecosystems

Cyber Twin Technology for AI-Driven Real-Time Software Secur...

引用

Mobile Networks and Wireless Communications (ICMNWC), IEEE International Conference on

作者： Krishna Chaitanya Sunkara Gulshan Dhasmana Ravi K S Nayana Yadav M Ishan Desai Rajmani Shankar Oracle Cloud Engineering Data Center Design & Modeling AL/ML Expert North Carolina USA Department of Computer Science & Engineering Graphic Era Deemed to be University Dehradun India Department of Mechanical Engineering Vidyavardhaka College of Engineering Mysuru India Department of Computer Science and Engineering AJ Institute of Engineering and Technology Mangalore India Department of Electrical Engineering Faculty of Engineering and Technology Parul institute of Engineering and Technology Parul University Gujarat India Faculty of Law Symbiosis Law School Symbiosis International (Deemed University) Pune India

ISBN: (数字)9798331528348

ISBN: (纸本)9798331528355

As Internet of Things (IoT) ecosystems grow more complex, ensuring real-time security has become a major challenge. Traditional security approaches are insufficient for handling dynamic and interconnected IoT networks, which are increasingly targeted by sophisticated cyber-attacks. To address these issues, new methodologies that combine real-time monitoring and adaptive security mechanisms are needed. Cyber Twin technology, an innovative extension of digital twin technology, presents a promising solution by creating AI-driven digital replicas of IoT devices and software systems for continuous security monitoring and management. This paper introduces a Cyber Twin Technology Framework for AI-driven real-time software security in IoT ecosystems. The framework employs advanced AI models, including Convolutional Neural Networks (CNNs) for anomaly detection and Generative Adversarial Networks (GANs) for synthetic data generation to simulate potential attack scenarios. A dynamic reinforcement learning module is integrated to optimize threat response strategies based on evolving threat patterns. By creating real-time digital replicas of IoT components, the Cyber Twin framework continuously monitors device behaviors, identifies anomalies, and autonomously initiates mitigation actions. The system is evaluated in a simulated IoT environment with over 500 interconnected devices. Experimental results demonstrate that the Cyber Twin framework achieved a 99.2% detection accuracy in identifying cyber threats, with a false positive rate of 1.3%. The dynamic response module reduced incident response time by 35% compared to traditional methods, enhancing the system's ability to neutralize potential threats in real-time. The use of GAN-based synthetic data also enabled proactive defense strategies, reducing attack success rates by 40% during testing. The Cyber Twin Technology Framework provides a robust solution for real-time software security in complex IoT ecosystems. By leveraging A

关键词： Accuracy Ecosystems Software systems Real-time systems Threat assessment Internet of Things Security Time factors Monitoring Synthetic data

来源：评论

学校读者我要写书评

暂无评论

UCPM: Uncertainty-Guided Cross-Modal Retrieval with Partially Mismatched Pairs

引用

IEEE Transactions on Image Processing 2025年 34卷 3622-3634页

作者： Zha, Quanxing Liu, Xin Cheung, Yiu-Ming Peng, Shu-Juan Xu, Xing Wang, Nannan Huaqiao University Department of Computer Science Xiamen 361021 China Key Laboratory of Pattern Recognition and Computer Vision Xiamen 361021 China Huaqiao University Fujian Key Laboratory of Big Data Intelligence and Security Xiamen 361021 China Hong Kong Baptist University Department of Computer Science Hong Kong Huaqiao University Department of Artificial Intelligence Xiamen 361021 China Fujian Province University Key Laboratory of Computer Vision and Machine Learning (Huaqiao University) Xiamen 361021 China University of Electronic Science and Technology of China Center for Future Multimedia School of Computer Science and Engineering Chengdu 610051 China Xidian University State Key Laboratory of Integrated Services Networks Xi’an 710071 China

The manual annotation of perfectly aligned labels for cross-modal retrieval (CMR) is incredibly labor-intensive. As an alternative, the collection of co-occurring data pairs from the Internet is a remarkably cost-effective way, but which, inevitably induces the Partially Mismatched Pairs (PMPs) and therefore significantly degrades the retrieval performance without particular treatment. Previous efforts often utilize the pair-wise similarity to filter out the mismatched pairs, and such operation is highly sensitive to mismatched or ambiguous data and thus leads to sub-optimal performance. To alleviate these concerns, we propose an efficient approach, termed UCPM, i.e., Uncertainty-guided Cross-modal retrieval with Partially Mismatched pairs, which can significantly reduce the adverse impact of mismatched data pairs. Specifically, a novel Uncertainty Guided Division (UGD) strategy is sophisticatedly designed to divide the corrupted training data into confident matched (clean), easily-identifiable mismatched (noisy) and hardly-determined hard subsets, and the derived uncertainty can simultaneously guide the informative pair learning while reducing the negative impact of potential mismatched pairs. Meanwhile, an effective Uncertainty Self-Correction (USC) mechanism is concurrently presented to accurately identify and rectify the fluctuated uncertainty during the training process, which further improves the stability and reliability of the estimated uncertainty. Besides, a Trusted Margin Loss (TML) is newly designed to enhance the discriminability between those hard pairs, by dynamically adjusting their soft margins to amplify the positive contributions of matched pairs while suppressing the negative impacts of mismatched pairs. Extensive experiments on three widely-used benchmark datasets, verify the effectiveness and reliability of UCPM compared with the existing SOTA approaches, and significantly improve the robustness in both synthetic and real-world PMPs. The code i

关键词： Cross-modal retrieval partially mismatched pairs trusted margin loss uncertainty guided division uncertainty self-correction

来源：评论

学校读者我要写书评

暂无评论

Motion-X++: A Large-Scale Multimodal 3D Whole-body Human Motion dataset

arXiv

引用

arXiv 2025年

作者： Zhang, Yuhong Lin, Jing Zeng, Ailing Wu, Guanlin Lu, Shunlin Fu, Yurong Cai, Yuanhao Zhang, Ruimao Wang, Haoqian Zhang, Lei Shenzhen International Graduate School Tsinghua University Guangdong 518000 China Department of Computer Vision and Robotics IDEA Guangdong 518000 China Whiting School of Engineering Johns Hopkins University MD21218 United States School of Data Science The Chinese University of Hong Kong Guangdong GD Shenzhen518000 China

In this paper, we introduce Motion-X++, a large-scale multimodal 3D expressive whole-body human motion dataset. Existing motion datasets predominantly capture body-only poses, lacking facial expressions, hand gestures, and fine-grained pose descriptions, and are typically limited to lab settings with manually labeled text descriptions, thereby restricting their scalability. To address this issue, we develop a scalable annotation pipeline that can automatically capture 3D whole-body human motion and comprehensive textural labels from RGB videos and build the Motion-X dataset comprising 81.1K text-motion pairs. Furthermore, we extend Motion-X into Motion-X++ by improving the annotation pipeline, introducing more data modalities, and scaling up the data quantities. Motion-X++ provides 19.5M 3D whole-body pose annotations covering 120.5K motion sequences from massive scenes, 80.8K RGB videos, 45.3K audios, 19.5M frame-level whole-body pose descriptions, and 120.5K sequence-level semantic labels. Comprehensive experiments validate the accuracy of our annotation pipeline and highlight Motion-X++’s significant benefits for generating expressive, precise, and natural motion with paired multimodal labels supporting several downstream tasks, including text-driven whole-body motion generation, audio-driven motion generation, 3D whole-body human mesh recovery, and 2D whole-body keypoints estimation, etc. Copyright © 2025, The Authors. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Identifying Sources of Soil Pollution in Industrial Areas Using IoT data and Neural Networks

Identifying Sources of Soil Pollution in Industrial Areas Us...

引用

Intelligent Technologies (ACOIT), Asian Conference on

作者： Rajesh Kambattan Kovarasan Dr N V Shibu A. Sairam K. K. Manivannan M.V Ishwarya T. R. GaneshBabu Department of Computer Science and Engineering Vel Tech Rangarajan Dr. Sagunthala R & D Institute of Science and Technology Chennai Tamil Nadu India Department of Computing Technologies School of Computing SRM Institute of Science and Technology Kattankulathur Chennai Tamil Nadu India Department of Computer Science and Engineering Saveetha School of Engineering Saveetha Institute of Medical and Technical Sciences Chennai Tamil Nadu India Department of Mechatronics Engineering KCG College of Technology Chennai Tamil Nadu India Artificial Intelligence and Data science Department Agni College of Technology Chennai Tamil Nadu India Department of Electronics and Communication Engineering Muthayammal Engineering College (Autonomous) Namakkal Tamil Nadu India

ISBN: (数字)9798350374957

ISBN: (纸本)9798350374964

Quick and precise source identification is crucial for developing successful remediation solutions to address soil contamination in industrial locations, which presents serious dangers to both the environment and human health. To locate the origins of soil contamination, this research presents a new method that combines neural network models with data collected from the Internet of Things (IOT). To monitor soil type, pollution levels, and weather conditions in real-time, set up IoT devices. After that, advanced neural network algorithms are used to process and analyze the data to find connections and patterns that might indicate the origins of the pollution. Our findings show that this method is useful for identifying where soil pollution inside industrial zones is coming from. Achieve high accuracy in source identification by integrating IoT technology and machine learning approaches. This allows us to address environmental harm and protect public health via focused remediation activities. Policymakers, environmental scientists, and stakeholders engaged in sustainable development and pollution control may benefit greatly which highlight the potential of IoT data and neural networks in challenges difficult environmental concerns.

关键词： Technological innovation Pollution Environmentally friendly manufacturing techniques Neural networks Pollution control Soil Internet of Things Stakeholders Soil pollution Contamination

来源：评论

学校读者我要写书评

暂无评论

Target-Oriented WiFi Sensing for Respiratory Healthcare: from Indiscriminate Perception to In-Area Sensing

引用

IEEE Network 2024年

作者： Wang, Meng Huang, Jinyang Zhang, Xiang Liu, Zhi Li, Meng Zhao, Peng Yan, Huan Sun, Xiao Dong, Mianxiong Hefei University of Technology Anhui Province Key Laboratory of Affective Computing and Advanced Intelligence Machine School of Computer and Information Hefei230601 China University of Science and Technology of China CAS Key Laboratory of Electromagnetic Space Information Hefei230026 China The University of Electro-Communications Department of Computer and Network Engineering Tokyo1828585 Japan Guizhou Normal University School of Big Data and Computer Science Guiyang550025 China Muroran Institute of Technology Department of Sciences and Informatics 0508585 Japan

Driven by the vision of integrated sensing and communication (ISAC) toward 6G technology, the WiFi-based respiration sensing approach has emerged as a highly competitive candidate for advanced healthcare services. Nevertheless, the indiscriminate perception of all the moving objects within the sensing area raises challenges for system stability and robustness for real-world deployment, especially for the susceptibility to motion interference from other people. Meanwhile, thanks to the emerging in-area wireless sensing technologies, i.e., beamforming, the dynamic environment tolerance of these technologies gives crucial opportunities for target-oriented WiFi-based healthcare services. This article discusses the compelling intersection of in-area WiFi sensing and the future of target-oriented intelligent respiratory healthcare network services. We first present a novel WiFi-based healthcare-assisting framework tailored for inter-user motion interference-tolerable (IMIT) sensing. The detailed components of low-level signal preprocessing, target-direction signal extraction, and healthcare-related services are discussed. By utilizing CSI beamforming technology, which adjusts a directional beam toward the desired direction and adaptively places beam nulls in noisy directions, we can mitigate motion interference and achieve target-oriented healthcare sensing. Finally, we discuss open challenges and potential solutions for target-oriented WiFi healthcare sensing. © 1986-2012 IEEE.

关键词： Wireless local area networks (WLAN)

来源：评论

学校读者我要写书评

暂无评论

Traffic4cast at NeurIPS 2022 – Predict Dynamics along Graph Edges from Sparse Node data: Whole City Traffic and ETA from Stationary Vehicle Detectors 36

Traffic4cast at NeurIPS 2022 – Predict Dynamics along Graph...

引用

36th Annual Conference on Neural Information Processing Systems - Competition Track, NeurIPS 2022

作者： Neun, Moritz Eichenberger, Christian Martin, Henry Spanring, Markus Siripurapu, Rahul Springer, Daniel Deng, Leyan Wu, Chenwang Lian, Defu Zhou, Min Lumiste, Martin Ilie, Andrei Wu, Xinhua Lyu, Cheng Lu, Qing-Long Mahajan, Vishal Lu, Yichao Li, Jiezhang Li, Junjun Gong, Yue-Jiao Grötschla, Florian Mathys, Joël Wei, Ye Haitao, He Fang, Hui Malm, Kevin Tang, Fei Kopp, Michael Kreil, David Hochreiter, Sepp Vienna Austria Institute of Cartography and Geoinformation ETH Zurich Switzerland School of Data Science University of Science and Technology of China China Huawei Noah’s Ark Lab. China Bolt Technology Tallinn Estonia University of Bucharest Bucharest Romania Department of Civil and Environmental Engineering Northeastern University BostonMA United States Technical University of Munich Germany Layer 6 AI Toronto Canada School of Coumpute Science and Engineering South China University of Technology Guangzhou China ETH Zurich Switzerland Department of Computer Science Loughborough University Loughborough United Kingdom School of Architecture Building and Civil Engineering Loughborough University Loughborough United Kingdom HERE Technologies ChicagoIL United States Kaiko Zurich Switzerland Machine Learning Institute Johannes Kepler University Linz Austria

The global trends of urbanization and increased personal mobility force us to rethink the way we live and use urban space. The Traffic4cast competition series tackles this problem in a data-driven way, advancing the latest methods in machine learning for modeling complex spatial systems over time. In this edition, our dynamic road graph data combine information from road maps, 1012 probe data points, and stationary vehicle detectors in three cities over the span of two years. While stationary vehicle detectors are the most accurate way to capture traffic volume, they are only available in few locations. Traffic4cast 2022 explores models that have the ability to generalize loosely related temporal vertex data on just a few nodes to predict dynamic future traffic states on the edges of the entire road graph. In the core challenge, participants are invited to predict the likelihoods of three congestion classes derived from the speed levels in the GPS data for the entire road graph in three cities 15 min into the future. We only provide vehicle count data from spatially sparse stationary vehicle detectors in these three cities as model input for this task. The data are aggregated in 15 min time bins for one hour prior to the prediction time. For the extended challenge, participants are tasked to predict the average travel times on super-segments 15 min into the future – super-segments are longer sequences of road segments in the graph. The competition results provide an important advance in the prediction of complex city-wide traffic states just from publicly available sparse vehicle data and without the need for large amounts of real-time floating vehicle data. © 2023 M. Neun et al.

关键词： Travel time

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：