检索结果-内蒙古大学图书馆

Subclass consistency regularization for learning with noisy labels based on contrastive learning

NEUROCOMPUTING 2025年 614卷

作者： Sun, Xinkai Zhang, Sanguo Chinese Acad Sci Key Lab Big Data Min & Knowledge Management Beijing 100049 Peoples R China

A prominent effect of label noise on neural networks is the disruption of the consistency of predictions. While prior efforts primarily focused on predictions' consistency at the individual instance level, they often fell short of fully harnessing the consistency across multiple instances. This paper introduces subclass consistency regularization (SCCR) to maximize the potential of this collective consistency of predictions. SCCR mitigates the impact of label noise on neural networks by imposing constraints on the consistency of predictions within each subclass. However, constructing high-quality subclasses poses a formidable challenge, which we formulate as a special clustering problem. To efficiently establish these subclasses, we incorporate a clustering-based contrastive learning framework. Additionally, we introduce the Q-enhancing algorithm to tailor the contrastive learning framework, ensuring alignment with subclass construction. We conducted comprehensive experiments using benchmark datasets and real datasets to evaluate the effectiveness of our proposed method under various scenarios with differing noise rates. The results unequivocally demonstrate the enhancement in classification accuracy, especially in challenging high-noise settings. Moreover, the refined contrastive learning framework significantly elevates the quality of subclasses even in the presence of noise. Furthermore, we delve into the compatibility of contrastive learning and learning with noisy labels, using the projection head as an illustrative example. This investigation sheds light on an aspect that has hitherto been overlooked in prior research efforts.

关键词： Deep learning Image classification Noisy label Consistency regularization Contrastive learning

来源：评论

学校读者我要写书评

暂无评论

MMA: Video Reconstruction for Spike Camera Based on Multiscale Temporal Modeling and Fine-Grained Attention

引用

IEEE SIGNAL PROCESSING LETTERS 2025年 32卷 1291-1295页

作者： Alim, Dilmurat Yang, Chen Qing, Laiyun Li, Guorong Huang, Qingming Univ Chinese Acad Sci Sch Comp Sci & Technol Key Lab Big Data Min & Knowledge Management Beijing 100049 Peoples R China

This paper presents a Multiscale Temporal Correlation Learning with the Mamba-Fused Attention Model (MMA), an efficient and effective method for reconstructing a video clip from a spike stream. Spike cameras offer unique advantages for capturing rapid scene changes with high temporal resolution. A spike stream contains sufficient information for multiple image reconstructions. However, existing methods generate only a single image at a time for a given spike stream, which results in excessive redundant computations between consecutive frames when aiming at restoring a video clip, thereby increasing computational costs significantly. The proposed MMA addresses such challenges by constructing a spike-to-video model, directly producing an image sequence at a time. Specifically, we propose a U-shaped Multiscale Temporal Correlation Learning (MTCL) to fuse the features at different temporal resolutions for clear video reconstruction. At each scale, we introduce a Fine-Grained Attention (FGA) module for fine-spatial context modeling within a patch and a Mamba module for integrating features across patches. Adopting a lightweight U-shaped structure and fine-grained feature extraction at each level, our method reconstructs high-quality image sequences quickly. The experimental results show that the proposed MMA surpasses current state-of-the-art methods in image quality, computation cost, and model size.

关键词： Feature extraction Streaming media Computational modeling Decoding Cameras Training Image quality Computer architecture Image sequences Spike cameras image reconstruction transformer Mamba Mamba

来源：评论

学校读者我要写书评

暂无评论

A Real-Time Lightweight Perceptron for Cloud-Edge Collaborative Predictive Maintenance of Online Service Systems

引用

IEEE INTERNET OF THINGS JOURNAL 2025年第9期12卷 12640-12657页

作者： Zhang, Linzi Shi, Yong Wang, Donghan Commun Univ China Sch Econ & Management Beijing 100024 Peoples R China Chinese Acad Sci Key Lab Big Data Min & Knowledge Management Res Ctr Fictitious Econ & Data Sci Beijing 100190 Peoples R China

In the era of expanding cloud-edge online service systems, predictive maintenance (PdM) based on key performance indicators (KPIs), such as CPU utilization, response rate, and network bandwidth, is essential for system operational reliability and security. Traditional data-driven PdM approaches for cloud-edge services often monitor individual indicators in isolation, neglecting their interrelationships. Although convolutional modules have been utilized for collaborative representation learning of localized high-dimensional KPIs, the high-computational complexity of CNNs and the limited kernel size hinder their efficacy in modeling high-dimensional, time-sensitive KPIs. To address these challenges, we propose a lightweight multichannel multilayer perceptron (MCMLP) framework for collaborative PdM. This framework enables efficient pointwise manipulation to capture inner-indicator hidden patterns within KPIs sequences without relying on grid-based convolution. Our MCMLP enhances locality through a two-stage dual-channel feature extraction process that captures spatial and temporal dependencies, integrating inner-indicator and cross-timeline features. The proposed MCMLP can also be easily plugged into existing CNN-based detectors as a substitute, offering a computationally economical alternative. Empirical results in real-world KPI data from various cloud-edge service scenarios demonstrate that MCMLP significantly outperforms traditional CNN-based methods, with a roughly 22% reduction in training time at a 95% confidence level (alpha = 0.05). This study also provides actionable insights for deploying MCMLP in cloud-edge service systems.

关键词： Feature extraction Convolutional neural networks Predictive maintenance Predictive models Long short term memory Computer architecture Collaboration Cloud computing Computational modeling Biological system modeling Bayesian-based posterior test channel fusion perceptron key performance indicators (KPIs) online service systems predictive maintenance (PdM)

来源：评论

学校读者我要写书评

暂无评论

A general tail item representation enhancement framework for sequential recommendation

引用

Frontiers of Computer Science 2024年第6期18卷 137-148页

作者： Mingyue CHENG Qi LIU Wenyu ZHANG Zhiding LIU Hongke ZHAO Enhong CHEN Anhui Province Key Laboratory of Big Data Analysis and Application University of Science and Technology of China&State Key Laboratory of Cognitive IntelligenceHefei 230026China College of Management and Economics Tianjin UniversityTianjin 300072China

Recently advancements in deep learning models have significantly facilitated the development of sequential recommender systems(SRS).However,the current deep model structures are limited in their ability to learn high-quality embeddings with insufficient ***,highly skewed long-tail distribution is very common in recommender ***,in this paper,we focus on enhancing the representation of tail items to improve sequential recommendation *** empirical studies on benchmarks,we surprisingly observe that both the ranking performance and training procedure are greatly hindered by the poorly optimized tail item *** address this issue,we propose a sequential recommendation framework named TailRec that enables contextual information of tail item well-leveraged and greatly improves its corresponding *** the characteristics of the sequential recommendation task,the surrounding interaction records of each tail item are regarded as contextual information without leveraging any additional side *** approach allows for the mining of contextual information from cross-sequence behaviors to boost the performance of sequential *** a light contextual filtering component is plug-and-play for a series of SRS *** verify the effectiveness of the proposed TailRec,we conduct extensive experiments over several popular benchmark *** experimental results demonstrate that TailRec can greatly improve the recommendation results and speed up the training *** codes of our methods have been available.

关键词： sequential recommendation long-tail distribution training accelerating

来源：评论

学校读者我要写书评

暂无评论

Dual-stream coupling network with wavelet transform for cross-resolution person re-identification

引用

Journal of Systems Engineering and Electronics 2023年第3期34卷 682-695页

作者： SUN Rui YANG Zi ZHAO Zhenghui ZHANG Xudong Key Laboratory of Knowledge Engineering with Big Data(Ministry of Education) Hefei University of TechnologyHefei 230601China School of Computer and Information Hefei University of TechnologyHefei 230601China

Person re-identification is a prevalent technology deployed on intelligent *** have been remarkable achievements in person re-identification methods based on the assumption that all person images have a sufficiently high resolution,yet such models are not applicable to the open *** real world,the changing distance between pedestrians and the camera renders the resolution of pedestrians captured by the camera *** low-resolution(LR)images in the query set are matched with high-resolution(HR)images in the gallery set,it degrades the performance of the pedestrian matching task due to the absent pedestrian critical information in LR *** address the above issues,we present a dualstream coupling network with wavelet transform(DSCWT)for the cross-resolution person re-identification ***,we use the multi-resolution analysis principle of wavelet transform to separately process the low-frequency and high-frequency regions of LR images,which is applied to restore the lost detail information of LR ***,we devise a residual knowledge constrained loss function that transfers knowledge between the two streams of LR images and HR images for accessing pedestrian invariant features at various *** qualitative and quantitative experiments across four benchmark datasets verify the superiority of the proposed approach.

关键词： cross-resolution feature invariant learning person re-identification residual knowledge transfer wavelet transform

来源：评论

学校读者我要写书评

暂无评论

Supply-Demand risk assessment of urban flood resilience from the perspective of the ecosystem services: A case study in Nanjing, China

引用

ECOLOGICAL INDICATORS 2025年 173卷

作者： Zhang, Peng Xu, Xukan Yang, Wentong Li, Yiming Yao, Shengqi Hohai Univ Business Sch Nanjing 211100 Peoples R China Hohai Univ Inst Stat & Data Sci Nanjing 211100 Peoples R China Key Lab Ind Big Data Min & Knowledge Management Changzhou 213000 Peoples R China

Urban flooding has become a pressing challenge for many countries and regions. Meanwhile, ecosystem-based disaster risk reduction approaches have been recognized as a sustainable and effective strategy for managing flood risks. This study designed a conceptual framework for assessing the supply-demand risk of urban flood resilience (UFR) from the perspective of ecosystem services (ESs). Taking the city of Nanjing, China, as an example, the InVEST model and the multi-criteria comprehensive evaluation method were employed to quantify the supply of UFR provided by natural ecosystems and the demand for UFR from socio-economic systems. Additionally, based on UFR supply-demand evaluation indicators calculated for each subdistrict, the Self- Organizing Map (SOM) was used to cluster the subdistricts. Finally, UFR supply-demand matching was conducted on the subdistrict clusters, and different flood-risk levels were identified based on the supply-demand ratio. The results showed that high flood-risk subdistricts are mainly concentrated in central urban area, low flood-risk subdistricts are primarily in urban periphery, and subdistricts in urban-rural transitional zones exhibit medium flood risk. Statistical analysis revealed that this zonal pattern is closely related to land use types and the distribution of social resources. Therefore, this study provides a scientific basis for developing management strategies of urban flood prevention from the perspective of ESs.

关键词： Urban flood resilience Ecosystem services Self-Organizing Map Supply-demand matching Risk assessment

来源：评论

学校读者我要写书评

暂无评论

Digitizing Highways:Unleashing Potential Through data Element Circulation for Safety,Cost Efficiency,and Environmental Impact Optimization

引用

Journal of Systems Science and Information 2025年第1期13卷 23-60页

作者： Yanyan JIA Wen LONG Yingjie TIAN School of Economics and Management University of Chinese Academy of SciencesBeijing 100190China Research Center on Fictitious Economy and Data Science Chinese Academy of SciencesBeijing 100190China Key Laboratory of Big Data Mining and Knowledge Management Chinese Academy of SciencesBeijing 100190China

As the era of large-scale highway maintenance arrives,the maintenance strategies have transitioned to a holistic approach that prioritizes safety,economic feasibility,and environmental *** research introduces a multi-objective optimization model for highway maintenance that incorporates the interplay of decision-maker preferences across three key objectives:Highway safety performance,maintenance engineering cost,and carbon *** study employs a large-sample data analysis on a subset of the Lianhuo Highway network,which includes 2,842 pavement *** approach mitigates the impact of outliers,ensuring a substantial data buffer that fortifies the model’s capacity for generalization and bolsters its *** findings reveal a Pareto-optimal relationship among the three scrutinized variables.A particularly noteworthy observation is the M-shaped trajectory of carbon emissions,which initially rise,then decline,and ultimately rebound,contingent upon the selected maintenance ***,an examination of the relationship between maintenance costs and safety performance discloses a trend of diminishing marginal returns,illustrating that the incremental gains in safety performance attenuate as maintenance investment escalates.

关键词： data element circulation highway maintenance multi-objective optimization

来源：评论

学校读者我要写书评

暂无评论

Leaving None Behind: data-Free Domain Incremental Learning for Major Depressive Disorder Detection

引用

IEEE Transactions on Affective Computing 2024年第2期16卷 758-770页

作者： Chen, Tao Guo, Yanrong Hao, Shijie Hong, Richang Hefei University of Technology Key Laboratory of Knowledge Engineering with Big Data China Hefei University of Technology Ministry of Education and School of Computer Science and Information Engineering Hefei230009 China

While deep learning techniques have shown promising performance in the Major Depressive Disorder (MDD) detection task, they still face limitations in real-world scenarios. Specifically, given the data scarcity, some efforts have resorted to aggregating data from different domains to expand the data volume. However, their effectiveness is currently limited by the domain gap and data privacy. Additionally, the class imbalance issue is particularly severe in our application, leading to biased classifying performance accordingly. To address these challenges, we propose data-Free Domain Incremental Learning for the MDD detection (DIL-MDD) task, accommodating multiple feature distributions by only accessing well-trained models from previous domains and the data in the current domain. Specifically, DIL-MDD consists of two key modules: Adaptive Class-tailored Threshold Learning (ACTL) and data-Free Domain Alignment (DFDA). The first module measures the discrepancy between the outputs of two sequential domains, based on which we learn a class-tailored threshold adaptively. Building on this, we differentiate between samples that either exhibit similarities or dissimilarities with the previous domain, where this similar sample set is identified to investigate the feature distribution of the historical data. The second module imposes an alignment constraint to narrow the gap between these two sample sets, thereby exploring the expertise of the previous domain. To validate the effectiveness of the proposed method, we conduct extensive experiments on the public MDD datasets, i.e., DAIC-WOZ, MODMA, and CMDC. We also apply our method to another mental health condition, Autism Spectrum Disorder (ASD), to further demonstrate its applicability. Finally, the ablation studies validate the superiority of the proposed modules. © 2010-2012 IEEE.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Fine-Grained Cross-Modal Fusion Based Refinement for Text-to-Image Synthesis

引用

Chinese Journal of Electronics 2023年第6期32卷 1329-1340页

作者： SUN Haoran WANG Yang LIU Haipeng QIAN Biao Department of Computer Science and Information Engineering Hefei University of Technology Key Laboratory of Knowledge Engineering with Big Data Ministry of EducationHefei University of Technology

Text-to-image synthesis refers to generating visual-realistic and semantically consistent images from given textual descriptions. Previous approaches generate an initial low-resolution image and then refine it to be high-resolution. Despite the remarkable progress, these methods are limited in fully utilizing the given texts and could generate text-mismatched images, especially when the text description is complex. We propose a novel finegrained text-image fusion based generative adversarial networks(FF-GAN), which consists of two modules: Finegrained text-image fusion block(FF-Block) and global semantic refinement(GSR). The proposed FF-Block integrates an attention block and several convolution layers to effectively fuse the fine-grained word-context features into the corresponding visual features, in which the text information is fully used to refine the initial image with more details. And the GSR is proposed to improve the global semantic consistency between linguistic and visual features during the refinement process. Extensive experiments on CUB-200 and COCO datasets demonstrate the superiority of FF-GAN over other state-of-the-art approaches in generating images with semantic consistency to the given texts.

关键词： Visualization Fuses Convolution Semantics Linguistics Benchmark testing Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Layer-Wise Learning Rate Optimization for Task-Dependent Fine-Tuning of Pre-Trained Models: An Evolutionary Approach

引用

ACM Transactions on Evolutionary Learning and Optimization 2024年第4期4卷 1-23页

作者： Bu, Chenyang Liu, Yuxin Huang, Manzong Shao, Jianxuan Ji, Shengwei Luo, Wenjian Wu, Xindong Key Laboratory of Knowledge Engineering with Big Data Ministry of Education and School of Computer Science and Information Engineering Hefei University of Technology Hefei China School of Artificial Intelligence and Big Data Hefei University Hefei China Guangdong Provincial Key Laboratory of Novel Security Intelligence Technologies School of Computer Science and Technology Harbin Institute of Technology Shenzhen China

The superior performance of large-scale pre-Trained models, such as Bidirectional Encoder Representations from Transformers (BERT) and Generative Pre-Trained Transformer (GPT), has received increasing attention in both academic and industrial research and has become one of the current research hotspots. A pre-Trained model refers to a model trained on large-scale unlabeled data, whose purpose is to learn general language representation or features for fine-Tuning or transfer learning in subsequent tasks. After pre-Training is complete, a small amount of labeled data can be used to fine-Tune the model for a specific task or domain. This two-stage method of "pre-Training+fine-Tuning"has achieved advanced results in natural language processing (NLP) tasks. Despite widespread adoption, existing fixed fine-Tuning schemes that adapt well to one NLP task may perform inconsistently on other NLP tasks given that different tasks have different latent semantic structures. In this article, we explore the effectiveness of automatic fine-Tuning pattern search for layer-wise learning rates from an evolutionary optimization perspective. Our goal is to use evolutionary algorithms to search for better task-dependent fine-Tuning patterns for specific NLP tasks than typical fixed fine-Tuning patterns. Experimental results on two real-world language benchmarks and three advanced pre-Training language models show the effectiveness and generality of the proposed framework. © 2024 held by the owner/author(s).

关键词： Semantics

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：