检索结果-内蒙古大学图书馆

Research on facial expression recognition based on multimodal data fusion and neural network

International Journal of Wireless and Mobile Computing 2024年第1期27卷 47-55页

作者： Han, Yi Wang, Xubin Lu, Zhengyu School of Mechanical Science and Engineering Huazhong University of Science and Technology Hubei Wuhan China Anyang Institute of Technology Department of Computer Science and Information Engineering Henan Province Anyang China School of Computer Science and Information Engineering Shanghai Institute of Technology FengXian District Shanghai China

Facial expression recognition is a challenging task when neural network is applied to pattern recognition. Most of the current recognition research is based on single source facial data, which generally has the disadvantages of low accuracy and low robustness. In this paper, a neural network algorithm of facial expression recognition based on multimodal data fusion is proposed. The algorithm is based on the multimodal data, and it takes the facial image, the histogram of oriented gradient of the image and the facial landmarks as the input, and establishes Convolutional Neural Network (CNN) designed to extract features from facial image, neural network designed to extract features from facial Landmarks and Neural Network (LNN) designed to extract features from Histogram of gradient (HNN), three sub-neural networks to extract data features, using multimodal data feature fusion mechanism to improve the accuracy of facial expression recognition. Experimental results show that, the algorithm has a great improvement in accuracy, robustness and detection speed. Copyright © 2024 Inderscience Enterprises Ltd.

关键词： Data fusion

来源：评论

学校读者我要写书评

暂无评论

Drug-Target Interactions Prediction Based on Signed Heterogeneous Graph Neural Networks

引用

Chinese Journal of Electronics 2024年第1期33卷 231-244页

作者： Ming CHEN Yajian JIANG Xiujuan LEI Yi PAN Chunyan JI Wei JIANG College of Information Science and Engineering Hunan Normal University School of Computer Science Shaanxi Normal University Faculty of Computer Science and Control Engineering Shenzhen Institute of Advanced TechnologyChinese Academy of Sciences Computer Science Department BNU-HKBU United International College

Drug-target interactions(DTIs) prediction plays an important role in the process of drug *** computational methods treat it as a binary prediction problem, determining whether there are connections between drugs and targets while ignoring relational types information. Considering the positive or negative effects of DTIs will facilitate the study on comprehensive mechanisms of multiple drugs on a common target, in this work, we model DTIs on signed heterogeneous networks, through categorizing interaction patterns of DTIs and additionally extracting interactions within drug pairs and target protein pairs. We propose signed heterogeneous graph neural networks(SHGNNs), further put forward an end-to-end framework for signed DTIs prediction, called SHGNN-DTI,which not only adapts to signed bipartite networks, but also could naturally incorporate auxiliary information from drug-drug interactions(DDIs) and protein-protein interactions(PPIs). For the framework, we solve the message passing and aggregation problem on signed DTI networks, and consider different training modes on the whole networks consisting of DTIs, DDIs and PPIs. Experiments are conducted on two datasets extracted from Drug Bank and related databases, under different settings of initial inputs, embedding dimensions and training modes. The prediction results show excellent performance in terms of metric indicators, and the feasibility is further verified by the case study with two drugs on breast cancer.

关键词： Drug-target interactions Signed heterogeneous network Link sign prediction Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

UAV-supported intelligent truth discovery to achieve low-cost communications in mobile crowd sensing

引用

Digital Communications and Networks 2024年第4期10卷 837-852页

作者： Jing Bai Jinsong Gui Guosheng Huang Shaobo Zhang Anfeng Liu School of Computer Science and Engineering Central South UniversityChangSha410083China School of Information Science and Engineering Hunan First Normal UniversityChangsha410205China School of Computer Science and Engineering Hunan University of Science and TechnologyXiangTan411100China

Unmanned and aerial systems as interactors among different system components for communications,have opened up great opportunities for truth data discovery in Mobile Crowd Sensing(MCS)which has not been properly solved in the *** this paper,an Unmanned Aerial Vehicles-supported Intelligent Truth Discovery(UAV-ITD)scheme is proposed to obtain truth data at low-cost communications for *** main innovations of the UAV-ITD scheme are as follows:(1)UAV-ITD scheme takes the first step in employing UAV joint Deep Matrix Factorization(DMF)to discover truth data based on the trust mechanism for an information Elicitation Without Verification(IEWV)problem in MCS.(2)This paper introduces a truth data discovery scheme for the first time that only needs to collect a part of n data samples to infer the data of the entire network with high accuracy,which saves more communication costs than most previous data collection schemes,where they collect n or kn data ***,we conducted extensive experiments to evaluate the UAV-ITD *** results show that compared with previous schemes,our scheme can reduce estimated truth error by 52.25%–96.09%,increase the accuracy of workers’trust evaluation by 0.68–61.82 times,and save recruitment costs by 24.08%–54.15%in truth data discovery.

关键词： Unmanned aerial systems Trust computing Truth discovery Deep matrix factorization Low-cost communications

来源：评论

学校读者我要写书评

暂无评论

Cognitive Transformation in Personal IoT: Pioneering Intelligent Automation

引用

Cyber-Physical Systems 2025年第2期11卷 183-240页

作者： Gulzar, Bisma Ahmad Sofi, Shabir Sholla, Sahil Department of Information Technology National Institute of Technology Srinagar India Department of Computer Science Engineering Islamic University of Science & Technology Srinagar India

In recent years, IoT has transformed personal environments by integrating diverse smart devices. This paper presents an advanced IoT architecture that optimizes network infrastructure, focusing on the adoption of MQTT protocol and introducing Cognitive Smart Objects for managing personal IoT applications. These objects use Neural Networks to predict optimal actions based on user behavior patterns. A Continuous Learning mechanism enables real-time adaptation of the network to evolving user interactions. The study highlights the role of Cognitive Transformation in Personal IoT, driving intelligent automation and enhancing user experience. © 2024 Informa UK Limited, trading as Taylor & Francis Group.

关键词： Internet of Things (IoT) Personal Internet of Things (PIoT) Social Internet of Things (SIoT) IoT architecture network infrastructure

来源：评论

学校读者我要写书评

暂无评论

DGC-TnT: Enhancing Point Cloud Object Classification by Dynamic Graph Convolutions with Transformer in Transformer

引用

IEEE Access 2024年 12卷 111924-111931页

作者： Lin, Chien-Chou Chen, Po-Yu National Yunlin University of Science and Technology Department of Computer Science and Information Engineering Douliou64002 Taiwan

Recently, deep learning neural networks have been widely used in object classification. The process of object classification typically involves extracting features from the point cloud using neural networks and integrating these features into a global feature for recognizing the object. However, most existing methods perform well in global feature extraction but still fall short in capturing local details, or lack an attention mechanism to focus on informative feature signals. This paper presents DGC-TnT which combines dynamic graph convolution (DGC) structure with the Point-TnT module. Employing a two-stage TnT architecture, DGC-TnT utilizes a self-attention mechanism to deeply analyze local features derived from dynamic graph convolution, expanding their scope globally. This mechanism effectively operates between individual points and clusters of points, thereby enhancing overall object classification performance. DGC-TnT facilitates enhanced feature communication between points and point clusters within the point cloud through the Transformer module. On the ModelNet40 and ScanobjectNN datasets, DGC-TnT has a significant improvement in accuracy. © 2013 IEEE.

关键词： Three dimensional displays

来源：评论

学校读者我要写书评

暂无评论

Color Image Compression and Encryption Algorithm Based on 2D Compressed Sensing and Hyperchaotic System

引用

computers, Materials & Continua 2024年第2期78卷 1977-1993页

作者： Zhiqing Dong Zhao Zhang Hongyan Zhou Xuebo Chen School of Computer Science and Software Engineering University of Science and Technology LiaoningAnshan114051China School of Electronic and Information Engineering University of Science and Technology LiaoningAnshan114051China

With the advent of the information security era,it is necessary to guarantee the privacy,accuracy,and dependable transfer of *** study presents a new approach to the encryption and compression of color *** is predicated on 2D compressed sensing(CS)and the hyperchaotic ***,an optimized Arnold scrambling algorithm is applied to the initial color images to ensure strong ***,the processed images are con-currently encrypted and compressed using 2D *** them,chaotic sequences replace traditional random measurement matrices to increase the system’s ***,the processed images are re-encrypted using a combination of permutation and diffusion *** addition,the 2D projected gradient with an embedding decryption(2DPG-ED)algorithm is used to reconstruct *** with the traditional reconstruction algorithm,the 2DPG-ED algorithm can improve security and reduce computational ***,it has better *** experimental outcome and the performance analysis indicate that this algorithm can withstand malicious attacks and prove the method is effective.

关键词： Image encryption image compression hyperchaotic system compressed sensing

来源：评论

学校读者我要写书评

暂无评论

A Multi-Agent Proximal Policy Optimized Joint Mechanism in mmWave HetNets with CoMP Toward Energy Efficiency Maximization

IEEE Transactions on Green Communications and Networking

引用

IEEE Transactions on Green Communications and Networking 2024年第1期8卷 265-278页

作者： Lotfolahi, Amin Ferng, Huei-Wen National Taiwan University of Science and Technology Department of Computer Science and Information Engineering Taipei106319 Taiwan

A novel cluster-based traffic offloading and user association (UA) algorithm alongside a multi-agent deep reinforcement learning (DRL) based base station (BS) activation mechanism is proposed in this paper. Our design aims to maximize the energy efficiency (EE) of the heterogeneous network (HetNet) while maintaining high quality of service (QoS). By taking advantage of the dense deployment of BSs in a HetNet, a clustering algorithm is first proposed to facilitate traffic offloading among BSs. Then, a multi-agent proximal policy optimization (MAPPO) based DRL algorithm is employed to trigger the BS activation decision based on the current environmental condition. Finally, a UA algorithm is deployed to improve further the (normalized) data rate of all users, known as the (normalized) sum rate. Via simulation, we show that our proposed mechanism can remarkably enhance the EE and excel over the closely related mechanisms. It satisfies the required data rate, improves the sum rate, and exhibits excellent scalability when many BSs are deployed. © 2017 IEEE.

关键词： Quality of service

来源：评论

学校读者我要写书评

暂无评论

Prognostic Survival Analysis for AD Diagnosis and Progression Using MRI Data: An AI-Based Approach

引用

IEEE Access 2025年 13卷 89059-89078页

作者： Deenadayalan, T. Shantharajah, S.P. Vellore Institute of Technology School of Computer Science Engineering and Information Systems Vellore632014 India

Alzheimer’s Disease is a progressive neuro-degenerative disorder and a leading cause of dementia, marked by cognitive decline, memory loss, and behavioral changes. Despite advancements in medical imaging and Artificial Intelligence (AI), early detection and accurate prognostic modeling remain challenging, particularly in distinguishing overlapping disease stages. This study proposes a novel AI framework combining EfficientNetB0, a state-of-the-art deep learning model, with dual attention mechanisms for robust feature extraction and survival analysis using the Alzheimer’s MRI dataset hosted on the Kaggle platform. The model was trained and evaluated on a balanced dataset of 6,400 MRI images, classified into 4 AD stages: Non-Demented (NoD), very Mild-Demented (vMiD), Mild-Demented (MiD), and Moderate-Demented (MoD). The pre-processing steps, including resizing, normalization, and data augmentation, ensured robustness and generalizability. The framework achieved classification accuracies of 99. 93% (training), 93. 60% (validation) and 93. 59% (testing), with a perfect Concordance Index of 1.0 for survival predictions, demonstrating its ability to accurately rank survival times. By integrating survival analysis with deep learning, the framework addresses limitations in previous work, such as reliance on linear assumptions and limited classification capabilities, and provides actionable insights for clinical workflows, supporting personalized treatment and early intervention strategies. Future work will focus on integrating longitudinal data, exploring time-dependent co-variates, and enhancing explainability to promote adoption in diverse clinical settings. This approach represents a significant advancement in AI-driven diagnosis and progression modeling of AD, with the potential to transform patient care and outcomes. © 2013 IEEE.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Hyperspectral Target Detection-Based 2-D-3-D Parallel Convolutional Neural Networks for Hyperspectral Image Classification

引用

IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 2024年 17卷 9451-9469页

作者： Chen, Shih-Yu Hsu, Kai-Hsun Kuo, Tzu-Hsien National Yunlin University of Science and Technology Department of Computer Science and Information Engineering Yunlin64002 Taiwan

This article presents a novel hyperspectral target detection (HTD) based two-dimensional (2-D)-three-dimensional (3-D) parallel convolutional neural network (HTD 2D-3D-PCNN) model, which integrates the HTD technique to achieve outstanding performance in hyperspectral image classification. The proposed model effectively leverages both spectral and spatial information through a dual-branch architecture in hyperspectral imaging. In the first branch, HTD is utilized to enhance the spectral features of targets of interest, while suppressing background. The resulting enhanced image is then inputted to a 2D-CNN, augmented with an additional deconvolution layer to highlight spatial characteristics. Concurrently, the second branch employs dimensionality reduction via principal component analysis, and a 3D-CNN is employed to capture both spectral and spatial attributes. Subsequently, the feature maps from both convolutional neural networks (CNNs) are combined and processed through fully connected layers for classification. To validate the effectiveness of the proposed HTD-2D-3D-PCNN, extensive experiments are conducted on five widely used hyperspectral public datasets (Indian Pines, Pavia University, Salinas Scene, Kennedy Space Center, and Botswana) with a consistent training sample ratio of either 10% or 5%. The results show that HTD-2D-3D-PCNN can achieve overall accuracy values of 98.41%, 99.85%, 99.92%, 99.82%, and 98.82% for the respective datasets, surpassing the performance of recent methodologies. © 2008-2012 IEEE.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

Solving Sparse Reward Tasks Using Self-Balancing Exploration and Exploitation

引用

Journal of Internet technology 2025年第3期26卷 293-301页

作者： Kong, Yan Wei, Junfeng Hsia, Chih-Hsien School of Computer and Software Nanjing University of Information Science and Technology China Department of Computer Science and Information Engineering National Ilan University Taiwan Department of Business Administration Chaoyang University of Technology Taiwan

A core challenge in applying deep reinforcement learning (DRL) to real-world tasks is the sparse reward problem, and shaping reward has been one effective method to solve it. However, due to the enormous state space and sparse rewards in the real world, a large number of useless samples may be generated, leading to reduced sample efficiency and potential local optima. To address this issue, this study proposes a self-balancing method of exploration and development to solve the issue of sparse rewards. Firstly, we shape the reward function according to the evaluated progress, to guide the agent’s learning of high-reward samples. Secondly, we construct a dual-trajectory exploration network, which provides intrinsic rewards based on the novelty of states and the trajectory difference of sibling agents to encourage the agent to explore and adjust the balance between exploration and exploitation. This method effectively prevents the generation of a large amount of useless training data during the interaction between the agent and the environment, resolves local optimal dilemmas through state novelty, and adjusts the strategy in a timely manner to solve sparse reward tasks. Our method outperforms basic reinforcement learning (RL) and curiosity-driven incentives in these experimental tasks. The self-balancing exploration and exploitation approach in our research provides a new perspective and effective solution for addressing the problem of sparse rewards, thereby advancing the application of DRL in real-world problems and achieving greater success. © 2025 Taiwan Academic Network Management Committee. All rights reserved.

关键词： Deep learning

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：