检索结果-内蒙古大学图书馆

GDKansformer: A group-wise dynamic Kolmogorov–Arnold transformer with multi-view gated attention for pathological image diagnosis

引用

Expert Systems with applications 2025年 286卷

作者： Lu, Xiaoyan Gong, Xun Chen, Yuanyuan School of Computing and Artificial Intelligence Southwest Jiaotong University Chengdu611756 China National Engineering Laboratory of Integrated Transportation Big Data Application Technology and Manufacturing Industry Chains Collaboration and Information Support Technology Key Laboratory of Sichuan Province Southwest Jiaotong University Chengdu611756 China Engineering Research Center of Sustainable Urban Intelligent Transportation Ministry of Education Chengdu611756 China Department of Pathology The Third People's Hospital of Chengdu Affiliated Hospital of Southwest Jiaotong University Chengdu610031 China

Pathological image diagnosis is a fundamental and critical component of precision medicine, playing a pivotal role in clinical decision-making. However, due to the high heterogeneity of pathological images, existing models still face the challenges of insufficient feature representation, poor interpretability, and high computational costs. To address these issues, we propose a novel pathological image diagnosis model called GDKansformer, which integrates the Group-wise Dynamic Kolmogorov–Arnold Network with a Vision Transformer (ViT) architecture. Specifically, this model employs adaptive dynamic fitting and group-wise computation strategies to seamlessly embed an improved Kolmogorov–Arnold Network (KAN) into ViTs, thereby boosting pathological representation and interpretability. The proposed group-wise computation strategy effectively mitigates the distribution overfitting caused by the KAN structure and significantly reduces the computational costs. Furthermore, a Multi-view Gated Attention (MVGA) block comprising pixel-level non-local neighborhoods and adaptive local perception is designed to eliminate redundant non-feature information and enhance contextual interactions, thereby improving the feature discrimination of the model for high heterogeneity data distributions. Experimental results demonstrate that GDKansformer achieves high accuracy with lower parameters, highlighting the significant potential in facilitating pathological diagnosis. © 2025

关键词： Pathology

来源：评论

学校读者我要写书评

暂无评论

Exploring Reliable PPG Authentication on Smartwatches in Daily Scenarios

arXiv

引用

arXiv 2025年

作者： Tang, Jiankai Liu, Jiacheng Tong, Renling Zhu, Kai Li, Zhe Yi, Xin Xing, Junliang Shi, Yuanchun Wang, Yuntao Key Laboratory of Pervasive Computing Ministry of Education Department of Computer Science and Technology Tsinghua University Ant Group China Key Laboratory of Pervasive Computing Ministry of Education Department of Computer Science and Technology Tsinghua University China Ant Group China Key Laboratory of Pervasive Computing Ministry of Education Department of Computer Science and Technology Tsinghua University China Intelligent Computing Application Laboratory of Qinghai Province Qinghai University China

Photoplethysmography (PPG) Sensors, widely deployed in smartwatches, offer a simple and non-invasive authentication approach for daily use. However, PPG authentication faces reliability issues due to motion artifacts from physical activity and physiological variability over time. To address these challenges, we propose MTL-RAPID, an efficient and reliable PPG authentication model, that employs a multitask joint training strategy, simultaneously assessing signal quality and verifying user identity. The joint optimization of these two tasks in MTL-RAPID results in a structure that outperforms models trained on individual tasks separately, achieving stronger performance with fewer parameters. In our comprehensive user studies regarding motion artifacts (N = 30), time variations (N = 32), and user preferences (N = 16), MTL-RAPID achieves a best AUC of 99.2% and an EER of 3.5%, outperforming existing baselines. We open-source our PPG authentication dataset along with the MTL-RAPID model to facilitate future research on GitHub. © 2025, CC BY.

关键词： Multi-task learning

来源：评论

学校读者我要写书评

暂无评论

Multimodal feature fusion based on object relation for video captioning

引用

CAAI Transactions on Intelligence Technology 2023年第1期8卷 247-259页

作者： Zhiwen Yan Ying Chen Jinlong Song Jia Zhu School of Computing Science South China Normal UniversityGuangzhouChina Key Laboratory of Intelligent Education Technology and Application of Zhejiang Province Zhejiang Normal UniversityJinhuaChina

Video captioning aims at automatically generating a natural language caption to describe the content of a ***,most of the existing methods in the video captioning task ignore the relationship between objects in the video and the correlation between multimodal features,and they also ignore the effect of caption length on the *** study proposes a novel video captioning framework(ORMF)based on the object relation graph and multimodal feature *** uses the similarity and Spatio-temporal relationship of objects in video to construct object relation features graph and introduce graph convolution network(GCN)to encode the object *** the same time,ORMF also constructs a multimodal features fusion network to learn the relationship between different modal *** multimodal feature fusion network is used to fuse the features of different ***,the proposed model calculates the length loss of the caption,making the caption get richer *** experimental results on two public datasets(Microsoft video captioning corpus[MSVD]and Microsoft research-video to text[MSR-VTT])demonstrate the effectiveness of our method.

关键词： approaches deep learning multimodel scene understanding video analysis

来源：评论

学校读者我要写书评

暂无评论

Gradient-Based Differential kWTA Network With application to Competitive Coordination of Multiple Robots

引用

IEEE/CAA Journal of Automatica Sinica 2022年第8期9卷 1452-1463页

作者： Mei Liu Xiaoyan Zhang Mingsheng Shang Long Jin School of Information Science and Engineering Lanzhou UniversityLanzhou 730000 State Key Laboratory of Tibetan Intelligent Information Processing and Application Xining 810016China Chongqing Key Laboratory of Big Data and Intelligent Computing Chongqing Institute of Green and Intelligent TechnologyChinese Academy of SciencesChongqing 400714China Shenzhen Research Institute Lanzhou UniversityShenzhen 518063China

Aiming at the k-winners-take-all(kWTA)operation,this paper proposes a gradient-based differential kWTA(GDk WTA)*** obtaining the network,theorems and related proofs are provided to guarantee the exponential convergence and noise resistance of the proposed GD-kWTA ***,numerical simulations are conducted to substantiate the preferable performance of the proposed network as compared with the traditional ***,the GD-k WTA network,backed with a consensus filter,is utilized as a robust control scheme for modeling the competition behavior in the multi-robot coordination,thereby further demonstrating its effectiveness and feasibility.

关键词： Consensus filter k-winners-take-all(kWTA) multirobot coordination noise resistance

来源：评论

学校读者我要写书评

暂无评论

Time-aware bike flow prediction framework with dynamic edge fusion and memory integration

引用

Knowledge-Based Systems 2025年 322卷

作者： M.Rajeh, Taha Li, Tianrui Luo, Zhpeng Javed, Muhammad Hafeez Alhaek, Fares School of Computing and Artificial Intelligence Southwest Jiaotong University Chengdu 611756 China Engineering Research Center of Sustainable Urban Intelligent Transportation Ministry of Education Chengdu 611756 China National Engineering Laboratory of Integrated Transportation Big Data Application Technology Southwest Jiaotong University Chengdu 611756 China Manufacturing Industry Chains Collaboration and Information Support Technology Key Laboratory of Sichuan Province Southwest Jiaotong University Chengdu 611756 China Stirling College Chengdu University Chengdu 610106 China Computing Science Faculty of Natural Sciences University of Stirling Stirling FK9 4LA United Kingdom

Bike-Sharing Systems (BSSs) have become a popular solution to urban sustainability concerns, providing affordable and environmentally friendly transportation. However, modeling and predicting shared-bike flow patterns are essential for developing intelligent transportation systems and facilitating city management. Despite the existence of several inflow and outflow prediction models, certain unresolved issues remain, including the inability to differentiate and capture heterogeneous dynamic relationships within daily flows, and the suboptimal utilization of historical flow data for forecasting. To address these limitations, we propose MVRC (Multi-View Relational Framework based on Clustering), a novel deep-learning framework for precise inflow and outflow prediction. The MVRC framework models three types of heterogeneous dynamic flow relationships during different periods of the day. Initially, we introduce a novel clustering method to group nearby stations exhibiting similar patterns, followed by creating a dynamic memory module to store and update edge attributes based on the time of day. Furthermore, the dynamic edge fusion mechanism effectively integrates heterogeneous, time-evolving flow features, capturing their varying impacts. To further enhance the model's capabilities, an offline component that leverages historical data is incorporated, improving performance during testing. Extensive experiments conducted on real datasets from New York City and Washington, D.C., demonstrate the superior prediction efficiency and framework effectiveness of the MVRC framework compared to existing state-of-the-art techniques. These results highlight the practicality and efficacy of MVRC in terms of prediction accuracy and stability. © 2025

关键词： Attention Bike flow Clustering Data fusion Flow prediction Graph convolutional networks Urban computing

来源：评论

学校读者我要写书评

暂无评论

Multi-label feature selection via exploring reliable instance similarities

引用

Knowledge-Based Systems 2025年 322卷

作者： Cong Guo Xizhe Wang Changqin Huang Yi Wang Chengling Gao Xiaodi Huang Key Laboratory of Intelligent Education Technology and Application of Zhejiang Province Zhejiang Normal University Jinhua China College of Education Zhejiang University Hangzhou China School of Computing Mathematics and Engineering Charles Sturt University Albury Australia

Existing multi-label feature selection methods commonly employ graph regularization to formulate sparse regression objectives based on manifold learning. However, these approaches face two primary challenges. First, they typically rely on fixed similarities derived from the original feature space, which can be unreliable due to irrelevant features. Second, while pseudo-labeling incorporates supervised information, traditional methods often neglect the global similarity structure among instances. To overcome these issues, we propose a two-stage iterative learning method that progressively refines instance similarities, mitigating the influence of irrelevant features. In the first stage (W-stage), we introduce reconstructed labels that integrate both global instance similarities and supervised information, enhancing the model’s capability to capture richer similarity structures among data points. We also integrate instance and label graph regularizations, coupled with an l 2 , 1 -norm constraint, into the objective function to optimize feature weights. In the second stage (S-stage), dynamically updated instance similarities derived from the reduced-dimensional space guide subsequent iterations, facilitating continuous refinement. These two stages collectively form a unified framework that iteratively enhances instance similarities within the reduced feature space while selecting relevant features. Experimental evaluations on 11 datasets demonstrate that our algorithm consistently outperforms eight state-of-the-art methods across six evaluation metrics. Moreover, our method achieves the highest overall performance ranking and exceeds the second-best approach by at least 2.19 ranks for each evaluation metric.

关键词： Semi-supervised learning

来源：评论

学校读者我要写书评

暂无评论

Randomized Latent Factor Model for High-dimensional and Sparse Matrices from Industrial applications

引用

IEEE/CAA Journal of Automatica Sinica 2019年第1期6卷 131-141页

作者： Mingsheng Shang Xin Luo Zhigang Liu Jia Chen Ye Yuan MengChu Zhou the Chongqing Engineering Research Center of Big Data Application for Smart Cities and Chongqing Key Laboratory of Big Data and Intelligent ComputingChongqing Institute of Green and Intelligent TechnologyChinese Academy of Sciences the School of Computer Science and Engineering Beihang University the Department of Electrical and Computer Engineering New Jersey Institute of Technology

Latent factor(LF) models are highly effective in extracting useful knowledge from High-Dimensional and Sparse(HiDS) matrices which are commonly seen in various industrial applications. An LF model usually adopts iterative optimizers,which may consume many iterations to achieve a local optima,resulting in considerable time cost. Hence, determining how to accelerate the training process for LF models has become a significant issue. To address this, this work proposes a randomized latent factor(RLF) model. It incorporates the principle of randomized learning techniques from neural networks into the LF analysis of HiDS matrices, thereby greatly alleviating computational burden. It also extends a standard learning process for randomized neural networks in context of LF analysis to make the resulting model represent an HiDS matrix *** results on three HiDS matrices from industrial applications demonstrate that compared with state-of-the-art LF models, RLF is able to achieve significantly higher computational efficiency and comparable prediction accuracy for missing data.I provides an important alternative approach to LF analysis of HiDS matrices, which is especially desired for industrial applications demanding highly efficient models.

关键词： Big data high-dimensional and sparse matrix latent factor analysis latent factor model randomized learning

来源：评论

学校读者我要写书评

暂无评论

Adaptive time-stepping and maximum-principle preserving Lagrangian schemes for gradient flows

arXiv

引用

arXiv 2025年

作者： Liu, Qianqian Chen, Wenbin Shen, Jie Cheng, Qing School of Mathematical Sciences Fudan University Shanghai200433 China Shanghai Key Laboratory for Contemporary Applied Mathematics School of Mathematical Sciences Fudan University Shanghai200433 China Eastern Institute of Technology Zhejiang Ningbo315200 China Key Laboratory of Intelligent Computing and Application Tongji University Ministry of Education Department of Mathematics Tongji University Shanghai200092 China

We develop in this paper an adaptive time-stepping approach for gradient flows with distinct treatments for conservative and non-conservative dynamics. For the non-conservative gradient flows in Lagrangian coordinates, we propose a modified formulation augmented by auxiliary terms to guarantee positivity of the determinant, and prove that the corresponding adaptive second-order Backward Difference Formulas (BDF2) scheme preserves energy stability and the maximum principle under the time-step ratio constraint 0 n ≤ rmax ≤ 32. On the other hand, for the conservative Wasserstein gradient flows in Lagrangian coordinates, we propose an adaptive BDF2 scheme which is shown to be energy dissipative, and positivity preserving under the time-step ratio constraint 0 n ≤ rmax ≤ 3+√217 in 1D and 0 n ≤ rmax ≤ 54 in 2D, respectively. We also present ample numerical simulations in 1D and 2D to validate the efficiency and accuracy of the proposed *** Codes 35K55, 35K65, 65M06, 65M12 Copyright © 2025, The Authors. All rights reserved.

关键词： Lagrange multipliers

来源：评论

学校读者我要写书评

暂无评论

Research on Kiwifruit Detection Algorithm Based on Faster-YOLOv8n 5

Research on Kiwifruit Detection Algorithm Based on Faster-YO...

引用

5th International Conference on Artificial Intelligence and Computer Engineering, ICAICE 2024

作者： Zhao, Dan Wang, Jiaxing Shan, Wei College of Physics and Electronic Information Huaibei Normal University China Anhui Provincial Key Laboratory of Intelligent Computing and Application Huaibei China

ISBN: (纸本)9798331528911

In view of the complex structure of YOLOv8n model and the large amount of parameters and calculations of the existing kiwifruit detection algorithm, it is difficult to deploy in the terminal. This paper proposes a lightweight kiwifruit detection algorithm model Faster-YOLOv8n based on YOLOv8n. That is, all C2f modules in the YOLOv8n model are merged with the FasterNet lightweight network structure to form a new C2f-Faster module to reduce the consumption of computing resources. The experimental results show that the average detection accuracy mAP of the Faster-YOLOv8n algorithm on the self-made kiwifruit dataset collected in this paper is 96.9 %. Compared with the YOLOv8n algorithm, the model size is reduced by 32 %,the calculation amount is reduced by 22.2 %, the parameter quantity is reduced by 29.6 %, and the detection speed is increased by 26.7 %.Therefore, the detection method proposed in this paper reduces the consumption of computing resources and provides a theoretical reference for the lightweight detection of kiwifruit. © 2024 IEEE.

关键词： Macroinvertebrates

来源：评论

学校读者我要写书评

暂无评论

Research on intelligent traffic pedestrian detection algorithm based on improved YOLOv8 24

Research on intelligent traffic pedestrian detection algorit...

引用

2024 2nd International Conference on Internet of Things and Cloud computing Technology, IoTCCT 2024

作者： Yue, Yurong Pan, Lihui Shan, Wei School of Physics and Electronic Information Huaibei Normal University Huaibei China Anhui Provincial Key Laboratory of Intelligent Computing and Application Huaibei China

ISBN: (纸本)9798400710148

Aiming at the problem that the pedestrian detection effect of illumination change, occlusion and different postures is not good in intelligent traffic scene, an optimized YOLOv8 is proposed to enhance its pedestrian detection ability. Based on YOLOv8, Coordinate Attention (CoordAtt) is introduced into the backbone network to enhance the sensitivity of the model to pedestrian location information and improve the detection accuracy. On the public test benchmark of Caltech Pedestrian Dataset, compared with the baseline algorithm, the proposed algorithm improves mAP0.5 by 1.3%, mAP0.5: 0.95 by 0.3%, and LAMR by 1.8%. Compared with the popular algorithm, mAP0.5 and mAP0.5: 0.95 are as high as 0.875 and 0.543 respectively, and LAMR is as low as 0.286, which is better than the popular algorithm. The algorithm has been tested to show that the accuracy of excellent performance detection has been significantly improved, and the phenomenon of missed detection has been greatly reduced, which can serve the intelligent transportation more effectively and reduce the potential safety hazards caused by missed detection. © 2024 Copyright held by the owner/author(s).

关键词： CoordAtt

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：