检索结果-内蒙古大学图书馆

2023 International Conference on Intelligent Systems, Communications, and Computer Networks, ISCCN 2023

作者： Wang, Qifa Shen, Qiwei State Key Laboratory Of Networking And Switching Technology China

ISBN: (纸本)9781510666238

Anomalies are rare items that differ significantly from the majority of the data and raise suspicion. Time series anomaly detection is of great significance in industrial applications, data mining and other fields. In this paper, we propose an autoencoder-based anomaly detection model. In the encoder part, the dependency relationship between different time series is mined through the graph attention network, the time domain data features are extracted through the GRU, and the frequency domain features are extracted through the convolutional neural network. The temporal data is reconstructed using a decoder consisting of a recurrent neural network in the encoder part. The residual between the reconstructed data and the original data is used to further judge anomalies. Data cleaning is performed in the preprocessing part to improve model performance. The effectiveness of our method is demonstrated on three publicly available datasets, and our method is found to outperform four other common anomaly detection methods. © 2023 SPIE. All rights reserved.

关键词： Anomaly detection

来源：评论

学校读者我要写书评

暂无评论

Knowledge Distillation for Image Restoration: Simultaneous Learning from Degraded and Clean Images

Knowledge Distillation for Image Restoration: Simultaneous L...

引用

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Zhang, Yongheng Yan, Danfeng State Key Laboratory of Networking and Switching Technology BUPT Beijing China

ISBN: (纸本)9798350368741

Model compression through knowledge distillation has seen extensive application in classification and segmentation tasks. However, its potential in image-to-image translation, particularly in image restoration, remains underexplored. To address this gap, we propose a Simultaneous Learning Knowledge Distillation (SLKD) framework tailored for model compression in image restoration tasks. SLKD employs a dual-teacher, single-student architecture with two distinct learning strategies: Degradation Removal Learning (DRL) and Image Reconstruction Learning (IRL), simultaneously. In DRL, the student encoder learns from Teacher A to focus on removing degradation factors, guided by a novel BRISQUE extractor. In IRL, the student decoder learns from Teacher B to reconstruct clean images, with the assistance of a proposed PIQE extractor. These strategies enable the student to learn from degraded and clean images simultaneously, ensuring high-quality compression of image restoration models. Experimental results across five datasets and three tasks demonstrate that SLKD achieves substantial reductions in FLOPs and parameters, exceeding 80%, while maintaining strong image restoration performance. © 2025 IEEE.

关键词： feature extractor image restoration Knowledge distillation simultaneous learning

来源：评论

学校读者我要写书评

暂无评论

Two-Hop Delay-Aware Energy Efficiency Resource Allocation in Space-Air-Ground Integrated Smart Grid Network

引用

Computers, Materials & Continua 2025年第5期83卷 2429-2447页

作者： Qinghai Ou Min Yang Jingcai Kong Yang Yang Beijing FibrLink Communications Co.Ltd. Beijing100071China State Key Laboratory of Networking and Switching Technology Beijing University of Posts and TelecommunicationsBeijing100876China

The lack of communication infrastructure in remote regions presents significant obstacles to gathering data from smart power sensors(SPSs)in smart grid *** such cases,a space-air-ground integrated network serves as an effective emergency *** study addresses the challenge of optimizing the energy efficiency of data transmission fromSPSs to low Earth orbit(LEO)satellites through unmanned aerial vehicles(UAVs),considering both effective capacity and fronthaul link capacity *** to the non-convex nature of the problem,the objective function is reformulated,and a delay-aware energy-efficient power allocation and UAV trajectory design(DEPATD)algorithm is proposed as a two-loop *** the inner loop remains non-convex,the block coordinate descent(BCD)method is employed to decompose it into three subproblems:power allocation for SPSs,power allocation for UAVs,and UAV trajectory *** first two subproblems are solved using the Lagrangian dual method,while the third is addressed with the successive convex approximation(SCA)*** iteratively solving these subproblems,an efficient algorithm is developed to resolve the inner loop *** results demonstrate that the energy efficiency of the proposed DEPATD algorithm improves by 4.02% compared to the benchmark algorithm when the maximum transmission power of the SPSs increases from 0.1 to 0.45W.

关键词： Energy efficiency effective capacity delay requirement power allocation smart grid space-air-ground integrated network

来源：评论

学校读者我要写书评

暂无评论

Soft Knowledge Distillation with Multi-Dimensional Cross-Net Attention for Image Restoration Models Compression

Soft Knowledge Distillation with Multi-Dimensional Cross-Net...

引用

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Zhang, Yongheng Yan, Danfeng State Key Laboratory of Networking and Switching Technology BUPT Beijing China

ISBN: (纸本)9798350368741

Transformer-based encoder-decoder models have achieved remarkable success in image-to-image transfer tasks, particularly in image restoration. However, their high computational complexity-manifested in elevated FLOPs and parameter counts-limits their application in real-world scenarios. Existing knowledge distillation methods in image restoration typically employ lightweight student models that directly mimic the intermediate features and reconstruction results of the teacher, overlooking the implicit attention relationships between them. To address this, we propose a Soft Knowledge Distillation (SKD) strategy that incorporates a Multi-dimensional Cross-net Attention (MCA) mechanism for compressing image restoration models. This mechanism facilitates interaction between the student and teacher across both channel and spatial dimensions, enabling the student to implicitly learn the attention matrices. Additionally, we employ a Gaussian kernel function to measure the distance between student and teacher features in kernel space, ensuring stable and efficient feature learning. To further enhance the quality of reconstructed images, we replace the commonly used L1 or KL divergence loss with a contrastive learning loss at the image level. Experiments on three tasks-image deraining, deblurring, and denoising-demonstrate that our SKD strategy significantly reduces computational complexity while maintaining strong image restoration capabilities. © 2025 IEEE.

关键词： contrastive learning image restoration Knowledge distillation multi-dimensional cross-net attention

来源：评论

学校读者我要写书评

暂无评论

Revisiting the Generalization Ability of Deep Local Descriptors

IEEE Transactions on Intelligent Vehicles

引用

IEEE Transactions on Intelligent Vehicles 2024年 1-11页

作者： Wang, Wufan Zhang, Bo Gao, Hui Que, Xirong Wang, Wendong State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing China Beijing University of Posts and Telecommunications Beijing China

Recently, learned local descriptors based on convolutional neural networks have achieved promising performance on standard benchmarks, yet still trailed the generalization ability of hand-crafted descriptors. In this paper, we analyze the reasons and mitigate the problem by promoting the learning of cross-domain discriminative features. Specifically, it is discovered that a vast number of easy positives are fed to the triplet margin loss during training, which can overwhelm the learning of informative hard negatives. To deal with it, we propose the Adaptive Positive Weighting Loss (APWL), which reshapes the triplet margin loss with a hardness-aware scaling factor to down-weight the contribution of easy positives and focus the model on hard negatives for learning discriminative features. Moreover, the domain discrepancy frequently encountered in the real world is observed to collapse the feature space of the target domain and impair the discriminative ability of learned descriptors. Therefore, a Cross-domain Distribution Regulation (CDR) is introduced to unfold the collapsed feature space so that discriminative features can be retained across domains to further boost the generalization ability of learned descriptors. Extensive experimental results demonstrate that when trained with the proposed loss and regulation term, the learned descriptors can be empowered with stronger generalization ability and outperform state-of-the-art methods on standard benchmarks. IEEE

关键词： Benchmarking

来源：评论

学校读者我要写书评

暂无评论

Decoupled Mean Filtering for Calibrated MUSIC Spectrum with Phase-Rotation Precoding under Cooperative ISAC Systems – From Interference Suppression Perspective

引用

IEEE Transactions on Signal Processing 2025年 73卷 1877-1892页

作者： Liu, Xiqing Li, Jingwen Gong, Jialong Yin, Meiyu Liu, Yuanwei Peng, Mugen Beijing University of Posts and Telecommunications State Key Laboratory of Networking and Switching Technology China University of Hong Kong Department of Electrical and Electronic Engineering Hong Kong

Integrated sensing and communication (ISAC) represents a critical scenario in the sixth-generation (6G) mobile communication, requiring systems to deliver both excellent communication and high-accuracy sensing capabilities. The requirement for high-accuracy sensing presents a major concern in scenarios with a single ISAC base station (BS). While a multi-BS collaborative ISAC system offers substantial improvements in sensing accuracy, mutual interference among BSs introduces a critical issue for implementation. In this work, we model the mutual interference in the process of multi-BS collaborative sensing and investigate the impact of these interferences on sensing accuracy across different modulation schemes. Based on this, we present a decoupled mean filtering (DMF) algorithm to mitigate interference, which performs effectively at low signal-to-interference-plus-noise ratios (SINRs) but shows limited effectiveness at high SINRs. Consequently, we propose an enhanced DMF integrated with the phase rotation precoding (PRP) tailored to various modulation types. The simulation results demonstrate that the proposed DMF-PRP algorithm effectively suppresses the sensing mutual interference and exhibits superior sensing accuracy compared to the existing schemes. © 1991-2012 IEEE.

关键词： Wiener filtering

来源：评论

学校读者我要写书评

暂无评论

RecStream: Graph-aware Stream Management for Concurrent Recommendation Model Online Serving 31

RecStream: Graph-aware Stream Management for Concurrent Reco...

引用

31st International Conference on Computational Linguistics, COLING 2025

作者： Guo, Shuxi Qi, Qi Sun, Haifeng Liao, Jianxin Wang, Jingyu State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing China

ISBN: (纸本)9798891761971

Recommendation Models (RMs) are crucial for predicting user preferences and enhancing personalized experiences on large-scale platforms. As the application of recommendation models grows, optimizing their online serving performance has become a significant challenge. However, current serving systems perform poorly under highly concurrent scenarios. To address this, we introduce RecStream, a system designed to optimize stream configurations based on model characteristics for handling high concurrency requests. We employ a hybrid Graph Neural Network architecture to determine the best configurations for various RMs. Experimental results demonstrate that RecStream achieves significant performance improvements, reducing latency by up to 74%. ©2025 Association for Computational Linguistics.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

MicroOps: Rapid Microservice Data Simulation and AIOps Model Development Platform 31

MicroOps: Rapid Microservice Data Simulation and AIOps Model...

引用

31st IEEE International Conference on Software Analysis, Evolution and Reengineering, SANER 2024

作者： Li, Yuewei Wang, Zhigang Qi, Qi Jing, Yuhan Wu, Jinming Wu, Zhikang Lu, Yan Wang, Chengsen Wang, Xingyu Wang, Jingyu Beijing University of Posts and Telecommunications State Key Laboratory of Networking and Switching Technology China

ISBN: (纸本)9798350330663

Artificial Intelligence for IT Operations (AIOps) for microservice systems has attracted much attention in academia and industry, aiming to reduce the burden of operations developers and improve the reliability of microservices. However, due to mostly private datasets and unique data requirements of different studies, researchers are forced to invest considerable effort in tedious tasks such as data simulation and data collection, which prevents them from concentrating on model development. To tackle this dilemma, we introduce MicroOps, a microservice data simulation and AIOps model development platform. MicroOps provides full-process automation support for microservice AIOps research, with key roles for rapid dataset generation and intuitive model testing. Based on MicroOps, we release two multimodal datasets collected from two widely used microservice systems. A user survey is conducted on MicroOps, evaluating its usability and practicality through the System Usability Scale (SUS) and open-ended questions. The results show that both are highly positively rated. Platform: https://***/OpenNetAI/MicroOps. © 2024 IEEE.

关键词： Simulation platform

来源：评论

学校读者我要写书评

暂无评论

Differential Privacy Federated Learning Based on Adaptive Adjustment

引用

Computers, Materials & Continua 2025年第3期82卷 4777-4795页

作者： Yanjin Cheng Wenmin Li Sujuan Qin Tengfei Tu The State Key Laboratory of Networking and Switching Technology Beijing University of Posts and TelecommunicationsBeijing100876China The School of Cyberspace Security Beijing University of Posts and TelecommunicationsBeijing100876China

Federated learning effectively alleviates privacy and security issues raised by the development of artificial intelligence through a distributed training *** research has shown that attackers can compromise user privacy and security by stealing model ***,differential privacy is applied in federated learning to further address malicious ***,the addition of noise and the update clipping mechanism in differential privacy jointly limit the further development of federated learning in privacy protection and performance ***,we propose an adaptive adjusted differential privacy federated learning ***,a dynamic adaptive privacy budget allocation strategy is proposed,which flexibly adjusts the privacy budget within a given range based on the client’s data volume and training requirements,thereby alleviating the loss of privacy budget and the magnitude of model ***,a longitudinal clipping differential privacy strategy is proposed,which based on the differences in factors that affect parameter updates,uses sparse methods to trim local updates,thereby reducing the impact of privacy pruning steps on model *** two strategies work together to ensure user privacy while the effect of differential privacy on model accuracy is *** evaluate the effectiveness of our method,we conducted extensive experiments on benchmark datasets,and the results showed that our proposed method performed well in terms of performance and privacy protection.

关键词： Federated learning privacy protection differential privacy deep learning

来源：评论

学校读者我要写书评

暂无评论

Making Pre-trained Language Models Better Continual Few-Shot Relation Extractors 30

Making Pre-trained Language Models Better Continual Few-Shot...

引用

Joint 30th International Conference on Computational Linguistics and 14th International Conference on Language Resources and Evaluation, LREC-COLING 2024

作者： Ma, Shengkun Han, Jiale Liang, Yi Cheng, Bo State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications China

ISBN: (纸本)9782493814104

Continual Few-shot Relation Extraction (CFRE) is a practical problem that requires the model to continuously learn novel relations while avoiding forgetting old ones with few labeled training data. The primary challenges are catastrophic forgetting and overfitting. This paper harnesses prompt learning to explore the implicit capabilities of pre-trained language models to address the above two challenges, thereby making language models better continual few-shot relation extractors. Specifically, we propose a Contrastive Prompt Learning framework, which designs prompt representation to acquire more generalized knowledge that can be easily adapted to old and new categories, and margin-based contrastive learning to focus more on hard samples, therefore alleviating catastrophic forgetting and overfitting issues. To further remedy overfitting in low-resource scenarios, we introduce an effective memory augmentation strategy that employs well-crafted prompts to guide ChatGPT in generating diverse samples. Extensive experiments demonstrate that our method outperforms state-of-the-art methods by a large margin and significantly mitigates catastrophic forgetting and overfitting in low-resource scenarios. © 2024 ELRA Language Resource Association: CC BY-NC 4.0.

关键词： Computational linguistics

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：