检索结果-内蒙古大学图书馆

UDI-3DVT: A Novel Network for 3-D Reconstruction via Uncertainty-Driven Sampling and Incremental Visual-Tactile Fusion

引用

IEEE Sensors Journal 2025年第6期25卷 10358-10369页

作者： Zhang, Chenyang Zhang, Shaohua Li, Yao Chen, Wenbai Gao, Guowei Wu, Peiliang Yanshan University Key Laboratory for Computer Virtual Technology and System Integration of Hebei Province School of Information Science and Engineering Qinhuangdao066004 China Beijing Information Science and Technology University School of Automation Beijing100192 China

The fusion of visual and tactile senses allows robots to reconstruct and understand objects, aiding in downstream tasks such as object recognition and object grasping. However, most existing visual-tactile reconstruction methods typically overlook the role of visual point cloud processing in reconstruction, sampling points using generative-based methods rather than uncertainty-driven approaches. In addition, these methods only concatenate visual and tactile features, failing to fully leverage the complementary advantages of these two modalities, resulting in unsatisfactory reconstruction results. To address this issue, leveraging the open-source GelSight tactile sensor, this article proposes a novel method called UDI-3DVT, which reconstructs 3-D shapes by combining visual point clouds and multiple tactile images. Specifically, we explore the uncertainty of point clouds, performing additional sampling on high-uncertainty edge points, and propose an attention-based point cloud sampling method. Then, the framework leverages the efficiency of convolutional neural network (CNN) in image processing and the transformer's superiority in context modeling by using CNN for tactile features and transformer for visual features. A feature fusion module based on 1 × 1 convolution is designed to aggregate features from two branches. By gradually adding tactile features, we achieve incremental visual-tactile integration. Furthermore, since the contact-based perception of tactile sensors may cause slight deformations in objects, UDI-3DVT supports the reconstruction of both rigid and deformable objects. Extensive experiments demonstrate that our proposed framework outperforms current state-of-the-art (SOTA) methods, especially the Chamfer distance (CD) metric increased by 14.8%. © 2001-2012 IEEE.

关键词： Image reconstruction

来源：评论

学校读者我要写书评

暂无评论

Unsupervised Outlier Detection with Reinforced Noise Discriminator

引用

ACM Transactions on Intelligent systems and technology 2025年第2期16卷 1-26页

作者： Zhang, Zhongping Liu, Daoheng Zhu, Jinwei Wu, Youxi Hebei Key Laboratory of Computer Virtual Technology and System Integration School of Information Science and Engineering Yanshan University Qinhuangdao China School of Information Science and Engineering Yanshan University Qinhuangdao China School of Artificial Intelligence Hebei University of Technology Tianjin China

Outlier detection is one of the hot topics in the field of machine learning and data mining. At present, there are many kinds of outlier detection algorithms. The accuracies of traditional outlier detection algorithms are often affected by unique parameters, and an increase in the amount of data and the dimensions of the data can seriously affect their efficiency and effectiveness. Methods based on generative adversarial networks (GANs) can solve the above problems, but they are unacceptable since the model often collapses during the training period. In this article, to solve the problems of curse of dimensionality and model collapse, we propose a novel reinforced noise discriminator (RND) method for unsupervised outlier detection in tabular data. We consider outlier detection as a binary classification problem. Thus, we apply a learnable reinforced discriminator and generate a large number of potential outliers with a uniform distribution and potential outliers that are close to the original data that are used as a negative sample to train the discriminator, which learns the distribution of the original data to detect outliers. We empirically compare the proposed approach with ten state-of-the-art outlier detection methods on both synthetic and real-world tabular datasets. The experimental results show that RND outperforms its competitors in the majority of cases. The codes used to perform the experiments described in this article are available at https://***/urlhearts/r-n-d. © 2025 Copyright held by the owner/author(s). Publication rights licensed to ACM.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Energy-efficient computation offloading via deep reinforcement learning in mobility-aware multi-access edge computing systems with diverse users

引用

Expert systems with Applications 2025年 286卷

作者： Wu, Haixing Jin, Shunfu School of Information Science and Engineering Yanshan University Qinhuangdao066004 China Key Laboratory for Computer Virtual Technology and System Integration of Hebei Province Yanshan University Qinhuangdao066004 China

Multi-access Edge Computing (MEC) has been a promising solution that enables Internet of Things (IoT) devices to support computation-intensive applications by offloading some tasks to the network edge. However, most existing offloading methods in MEC systems are based on the premise of scenarios with either single-type or stable users, which is not aligned with the practical situations of users’ diversity and mobility. In addition, these methods always ignore energy saving strategies in MEC systems, which inevitably degrades Utility Energy Efficiency (UEE). To tackle these issues, considering personalized requirements from diverse users and idle energy consumption in MEC servers, we propose a novel computation offloading architecture with adaptive sleeping mechanism and heterogeneous MEC servers. Based on the architecture, we investigate energy-efficient computation offloading in mobility-aware MEC systems with diverse users. Under the Deep Reinforcement Learning (DRL) framework, we propose a Twin Delayed Deep Deterministic Policy Gradient Algorithm with Bursty Traffic (TD3-BT), in which the inherent correlation of task arrivals is taken into account to capture the behavior of correlated traffic and provide rational metrics estimation within one time slot. A long-term system cost minimization problem is formulated to optimize computation offloading for the trade-off between system delay and UEE. Assisted by TD3-BT, the formulated problem is effectively solved and the decision making in each time slot is provided for diverse users. Experimental results demonstrate that our TD3-BT algorithm achieves superior performance under various offloading scenarios compared to some benchmarks. © 2025 Elsevier Ltd

关键词： Energy efficiency

来源：评论

学校读者我要写书评

暂无评论

DSAA: Cross-modal transferable double sparse adversarial attacks from images to videos

引用

Neurocomputing 2025年 639卷

作者： Han, Xueqiang Zhang, Shihui Wang, Houlin Tian, Qing School of Information Science and Engineering Yanshan University Hebei Qinhuangdao China The Key Laboratory for Computer Virtual Technology and System Integration of Hebei Province Hebei Qinhuangdao China Department of Computer Science University of Alabama at Birmingham AL United States

Recent preliminary research has examined cross-modal attacks, enabling the transition of unlabeled attacks from images to videos. The current cross-modal attacks are mainly dense attacks, while sparse attacks have not been explored. To bridge this gap, this paper proposes a cross-modal transferable double sparse adversarial attack (DSAA) method. Specifically, DSAA incorporates both spatial sparsity and temporal sparsity. Firstly, we design a novel double-cross-shaped patch paradigm to achieve spatial sparsity. Leveraging this patch for perturbation addition enables a more effective global impact on video frames. Subsequently, we propose a uniform temporal sparsity strategy to achieve temporal sparsity. Even if only half of the frames are perturbed, the resulting average attack success rate remains comparable to that of perturbing all frames. Furthermore, we take into account the luminance, contrast, and structure of images and propose a structural similarity loss function (SSLF). By utilizing SSLF to measure the similarity of intermediate feature maps and incorporating it into perturbation optimization, DSAA achieves excellent attack performance. Extensive experiments demonstrate DSAA's superiority over state-of-the-art approaches in terms of attack success rate. The proposed DSAA uses video adversarial examples generated by four ImageNet-pretrained image models on Kinetics-400 and UCF-101 to attack six video recognition models, achieving average attack success rates of 85.83% and 82.02%, respectively. © 2025 Elsevier B.V.

关键词： Cross-modal attacks Spatial sparsity Temporal sparsity Unlabeled attacks Video adversarial examples

来源：评论

学校读者我要写书评

暂无评论

Towards self-interpretable review spammer group detection

引用

Information Sciences 2025年 716卷

作者： Huo, Chenghang Zhang, Fuzhi School of Information Science and Engineering Yanshan University Hebei Province Qinhuangdao China Hebei Key Laboratory of Computer Virtual Technology and System Integration Hebei Province Qinhuangdao China

Due to data imbalance, existing spammer group detection methods often yield suboptimal performance. Moreover, many of these approaches operate as black boxes, offering little to no interpretability for their detection results. To address these issues, we propose a self-interpretable approach for detecting review spammer groups. Our method begins by constructing a user-product bipartite graph with edge attributes. We integrate user review information with a novel fitness function to develop an adaptive genetic algorithm that effectively identifies high-quality candidate groups. Next, we introduce a hybrid graph neural network enhanced with active learning to generate vector representations of these candidate groups. We then design and construct a prototype layer and a group classification layer to detect spammer groups accurately. To provide interpretability, we incorporate prototype learning to create an interpretation mechanism that explains detection outcomes. Experimental results demonstrate that our method achieves substantial improvements in Precision@k and Recall@k at the top-1000 ranking, outperforming state-of-the-art solutions on the Amazon, YelpChi, YelpNYC, and YelpZip datasets by [11.53 %, 97.36 %, 51.37 %, 32.06 %] and [12.65 %, 54.65 %, 64.19 %, 47.37 %], respectively. Additionally, the Fidelity of our interpretability results under varying Sparsity levels is approximately 4 %, 9 %, 8 %, and 8 % higher than those of existing methods on the same datasets. © 2025 Elsevier Inc.

关键词： Active learning

来源：评论

学校读者我要写书评

暂无评论

Geo-Enhanced High-Order Feature Interaction Network for Cloud API QoS Prediction 19th

Geo-Enhanced High-Order Feature Interaction Network for Clou...

引用

19th CCF Conference on computer Supported Cooperative Work and Social Computing, ChineseCSCW 2024

作者： Ma, Xinhong Zeng, Yehong Qi, Mingyang Si, Yali Chen, Zhen School of Information Science and Engineering Yanshan University Qinhuangdao066004 China Key Laboratory for Computer Virtual Technology and System Integration of Hebei Province Yanshan University Qinhuangdao066004 China Changshu Institute of Technology Jiangsu Changshu215500 China

ISBN: (纸本)9789819623754

In the era of open cloud, cloud API is an important component and key enabling technology to achieve efficient data transmission, artificial intelligence algorithm empowerment, and software development cost reduction and efficiency. With the proliferation of cloud API with similar functionality, quality of service (QoS) becomes essential for differentiating cloud API performance, and QoS-driven prediction for cloud API has emerged as a critical topic for high quality cloud API selection. However, existing methods often inadequately utilize contextual information in resource-constrained scenario, limiting the improvement of cloud API prediction performance. We in this paper propose a geo-enhanced high-order feature interaction network (GHFIN) for cloud API QoS prediction. Firstly, a correlation analysis was conducted on real-world QoS dataset to explore the relationship between QoS data and geospatial context. Then, we design a geo-enhanced strategy that effectively utilizes the hierarchical relationships of geospatial entities (country, province, city, and district) through an offline real-time coordinate conversion module. On this basis, a geo-enhanced cloud API QoS prediction network GHFIN was designed, which aims to learn the high-order feature interactions through layer-wise feature stacking. Experimental results on two real-world QoS datasets illustrate the effectiveness of GHFIN compared to the classic and state-of-the-art methods, and our extensive analysis shows how the geospatial context and high-order feature interaction positively impact the performance of cloud API QoS prediction. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Quality of service

来源：评论

学校读者我要写书评

暂无评论

PromptHC: Multi-attention prompt guided haze-weather crowd counting

引用

Expert systems with Applications 2025年 286卷

作者： Shen, Jienan Tong, Liangang Li, Shaohua Kong, Weihang School of Information Science and Engineering Yanshan University Qinhuangdao066004 China Information Technology Center Yanshan University Qinhuangdao066004 China Hebei Key Laboratory of Computer Virtual Technology and System Integration Qinhuangdao066004 China

Existing crowd counting methods encounter the challenge of degraded performance in hazy weather due to the blurring of pedestrian outlines. However, current hazy-weather crowd counting methods primarily focus on extracting crowd features, often neglecting the varying degrees of distortion in pedestrian outlines caused by inhomogeneous haze distribution. To this end, this paper develops a multi-attention prompt guided method for hazy-weather crowd counting, termed PromptHC. Specially, to explore the relationship between varying haze concentrations and the pedestrian outlines, a multi-attention dynamically adjustable prompt module is designed to provide crucial prompts about crowd features in hazy weather. Meanwhile, to further enhance the anti-interference capability of the model in hazy weather, a progressive guidance module is incorporated, which effectively reduces interference from different haze concentrations by guiding the learning of crowd attention. Furthermore, a global context-enhanced crowd feature extraction module is designed to capture precise global information. A series of ablation studies verify the actual effectiveness of each core component of the PromptHC. In addition, we conduct a performance comparison with the current mainstream methods on two hazy-weather datasets. Experimental results show the feasibility and superiority of the PromptHC for the hazy-weather crowd counting task. © 2025 Elsevier Ltd

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Multi-level graph contrastive learning for cold-start recommendation in mashup development

引用

Information Sciences 2025年 717卷

作者： Sun, Mengmeng Xu, Yueshen Tan, Zeyu You, Dianlong Chen, Zhen School of Information Science and Engineering Yanshan University Qinhuangdao 066004 China School of Computer Science and Technology Xidian University Xi'an 710071 China Hebei Key Laboratory for Computer Virtual Technology and System Integration Yanshan University Qinhuangdao 066004 China

One longstanding challenge in cloud API recommender systems is the Mashup cold-start problem, i.e., to recommend suitable cloud APIs for new Mashups without any historical invocation records. To address this challenge, researchers typically integrate various auxiliary information and feed them into a supervised learning model, or provide contextual information beyond interaction behaviors via explicit graph connections. However, these researches fail to 1) explore the functionally complementary relationship between Mashups and cloud APIs, and 2) model their semantic associations. To this end, we propose a Multi-level Graph Contrastive Learning, namely MGCL. Unlike traditional single-view methods, we comprehensively model Mashup-API relationships from three views: interaction, functionally complementary, and semantics. The interaction view captures dynamic behaviors, while the other two reveal associations based on static attributes. Additionally, we design a multi-level graph contrastive learning mechanism to fuse information from these views in a self-supervised manner, enabling both cold- and warm-start Mashup representations to benefit from dynamic interaction behaviors and static attributes. Extensive experiments on real-world datasets demonstrate that MGCL outperforms state-of-the-art methods in both Mashup cold- and warm-start scenarios. The implementations are available at: https://***/MengMeng3399/MGCL. © 2025 Elsevier Inc.

关键词： Cloud API Cold start recommendation Functionally complementary Graph contrastive learning Mashup development Semantic association

来源：评论

学校读者我要写书评

暂无评论

Multi-label feature selection with feature reconstruction and label correlations

引用

Expert systems with Applications 2025年 285卷

作者： Zhang, Ning Wang, Ao Lu, Pengwei Feng, Tao Xu, Ying Du, Guodong School of Information Science and Engineering Yanshan University Qinhuangdao066004 China Hebei Key Laboratory of Computer Virtual Technology and System Integration Yanshan University Qinhuangdao066004 China Department of Computer Science and Technology Tsinghua University Beijing100084 China Shenzhen Research Institute of Yanshan University Shenzhen518063 China

With the arrival of the big data era, multi-label data have frequently appeared in various fields. However, multi-label data often contain rich feature information, leading to a high-dimensional feature space. Thus, multi-label feature selection plays a crucial role in addressing this challenge as a preprocessing step. Multi-label data assign a set of labels to each instance simultaneously;therefore, capturing the correlations among labels is an effective means of enhancing the learning performance. However, current multi-label feature selection methods seldom consider label correlations. Moreover, redundant features can cause overfitting, increase the complexity, and reduce the interpretability. To address these issues, we propose a multi-label feature selection method that incorporates feature reconstruction and label correlations. Specifically, we construct a low-dimensional embedding from the original feature space using induction, which preserves the manifold structure of the feature space. Next, we explore the global and local label correlations using a manifold regularization term designed by the generated low-dimensional space. We then achieve feature reconstruction by minimising the reconstruction error and sparse regularization terms to learn the relationships between the features. Finally, we construct an optimal objective function with 2,1-norm regularization to facilitate multi-label feature selection and ensure convergence. Comparative experiments on 11 datasets show that the proposed method exhibits notable performance improvements over existing methods(e.g., achieves 5.6 % higher Average precision than RFS on Scene and 3.6 % lower Coverage than CRASG on Science), validating the effectiveness and superiority of the proposed method. © 2025 Elsevier Ltd

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Depression Scale Dictionary Decomposition Framework for Multimodal Automatic Depression Level Prediction

引用

IEEE Transactions on Circuits and systems for Video technology 2025年

作者： Niu, Mingyue Wang, Xu Gong, Jibing Liu, Bin Tao, Jianhua Schuller, Bjorn W. Yanshan University School of Information Science and Engineering Hebei province Qinhuangdao066004 China Hebei Key Laboratory of Computer Virtual Technology and System Integration China Technology Co. Ltd Bejing100022 China Bejing100190 China Tsinghua University Department of Automation Bejing100190 China University of Chinese Academy of Sciences School of Artificial Intelligence Beijing100049 China University of Augsburg ZD.B. Chair of Embedded Intelligence for Health Care and Wellbeing Augsburg Germany Glam - Group on Language Audio and Music United Kingdom

Currently, many researchers aim to achieve automatic depression level prediction via speech and video behavior analysis. However, previous works have struggled to decompose audio and video sequences into the information related to and unrelated to depression scores, hindering the model's perception of depression cues. Besides, previous works implement multimodal fusion using attention mechanisms or linear layers, but failed to simultaneously consider the Euclidean relationship among tokens and the non-Euclidean relationship among channels, which bring limitations in capturing depression cues. In response to the above issues, we propose a depression scale dictionary decomposition framework, which mainly includes a Bidirectional Dictionary Decomposition (BDD) module and a Bidirectional Multimodal Fusion (BMF) module. The BDD module can use the dictionaries generated based on the depression scale to semantically decompose audio and video sequences into the information related to and unrelated to depression scores along token and channel dimensions for promoting depression cue perception. Moreover, considering the respective characteristics of tokens and channels, the BMF module uses linear layers and graph convolution to achieve cross-modal mixing, which is used to aggregate audio and video sequences for predicting depression levels. The validation on AVEC 2013, AVEC 2014 and DAIC-WOZ datasets demonstrates our method's superiority. © 1991-2012 IEEE.

关键词： Boolean functions

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

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

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：