检索结果-内蒙古大学图书馆

MSMFT: Multi-Stream Multimodal Factorized Transformer for Human Activity Recognition

IEEE Sensors Journal 2025年第6期25卷 10402-10416页

作者： Zhou, Xianfa Yuan, Jiabin Fan, Lili Niu, Xuewei Zha, Keke Liu, Xu Nanjing University of Aeronautics and Astronautics Department of Computer Science and Technology Nanjing210016 China Jinling Institute of Technology Department of Computer Engineering Nanjing211169 China

Existing video-based human activity recognition (HAR) methods are susceptible to challenges such as lighting variations and occlusions in complex environments. Wearable sensors can effectively mitigate these issues. The fusion of the two can reduce environmental impacts, thus enhancing accuracy. However, the heterogeneity between the two data types presents a challenge for fusion. To address this, we propose an architecture called multi-stream multimodal factorized transformer (MSMFT), which leverages the aggregation capabilities of transformers to achieve alignment and fusion of multimodal features without relying on specific feature extractors. First, we designed a statistical feature projection (SFP) method to effectively tokenize wearable sensor data for transformer models. Second, we introduced temporal embedding (TE) to capture local temporal feature variations. Then, we incorporated a factorized transformer architecture that includes temporal and spatial transformer encoders (STEs) to extract global spatiotemporal features fully. Finally, to avoid the issue of modality entanglement caused by deep fusion, we constructed a multi-stream architecture comprising a video stream, wearable sensors stream, and fusion stream. This design allows for the extraction of both independent and interactive modality features. The performance of the MSMFT method was evaluated on three multimodal datasets: UP-Fall, UTD-MHAD, and MMAct. The results show that MSMFT performs better than the state-of-the-art, with a 16.83% improvement in cross-scene F1-score on the MMAct dataset. © 2001-2012 IEEE.

关键词： Wearable sensors

来源：评论

学校读者我要写书评

暂无评论

Optimizing medical image analysis through MViTX on multiple datasets with explainable AI

引用

Neural Computing and Applications 2025年第14期37卷 8479-8507页

作者： Sheakh, Md. Alif Tahosin, Mst. Sazia Alam, Mohammad Jahangir Begum, Mahbuba Department of Computer Science and Engineering Daffodil International University Dhaka Bangladesh Department of Computer Science and Engineering Mawlana Bhashani Science and Technology University Tangail Bangladesh

Cancer remains a leading cause of mortality worldwide, with early detection and accurate diagnosis critical to improving patient outcomes. While computer-aided diagnosis systems powered by deep learning have shown considerable promise, their widespread clinical adoption faces significant challenges in maintaining consistent performance across diverse imaging modalities and datasets. This research addresses the critical challenge of developing robust, generalizable deep learning models by proposing a comprehensive framework utilizing seven diverse medical imaging datasets encompassing fundus photography, histopathology, endoscopy, and MRI, covering diseases such as ocular toxoplasmosis, endometrial cancer, colorectal cancer, gastrointestinal disease, breast cancer, brain tumor, and tympanic membrane conditions. Our methodology combines customized data augmentation strategies (photometric, geometric, and elastic transformations) with an optimized vision transformer with external attention (MViTX) architecture. The MViTX model demonstrated exceptional performance with test accuracies ranging from 94.1 to 99.1% across all datasets, achieving superior metrics in accuracy, precision, recall, F1-score, and AUC compared to state-of-the-art CNNs. The model's effectiveness was further validated through ablation studies and explainable AI techniques, while its practical utility was demonstrated through deployment as a user-friendly web application. Our research establishes the effectiveness of combining tailored data augmentation with attention-based transformer architectures for medical image analysis, representing a significant step toward enhancing healthcare professionals' diagnostic capabilities and ultimately improving patient care outcomes. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2025.

关键词： Magnetic resonance imaging

来源：评论

学校读者我要写书评

暂无评论

Enhancing patient-independent detection of freezing of gait in Parkinson’s disease with deep adversarial network

引用

Neural Computing and Applications 2025年 1-21页

作者： Fahad, Md Shah Ranjan, Ashish Kumar, Gautam Department of Computer Science and Engineering Birla Institute of Technology Mesra Jharkhand Ranchi India Department of Computer Science and Engineering C. V. Raman Global University Odisha Bhubaneswar India Department of Computer Science and Engineering National Institute of Technology Delhi Delhi India

Freezing of gait (FoG) refers to sudden, relatively brief episodes of gait arrest in Parkinson’s disease, known to manifest in the advanced stages of the condition. Events of freezing are associated with tumbles, traumas, and psychological repercussions, significantly impacting the patient’s quality of life. The use of accelerometer data derived from sensors put on a patient’s body to detect FoG has previously been proposed using convolutional neural network (CNN)-based deep learning algorithms. Here, we combine a CNN + Long short-term Memory (LSTM)-attention model to detect FoG episodes—a first for the detection of FoG using the accelerometer data. CNN facilitates automatic feature extraction from the accelerometer data itself. The output from the CNN is fed to the LSTM network, which is known for capturing sequential information. Further, the attention mechanism introduces relative focus on individual sub-sequences during the training of the LSTM network. The proposed model is made patient independent using adversarial training. The proposed model achieved improvements of +7.60% (without adversarial training) and + 8.36% (with adversarial training) in the sensitivity values over state of the art. This is obtained with little or no compromise in specificity values. The corresponding improvements in accuracy values are +3.81 and +2.23%, respectively. These findings imply that the proposed method can detect FoG gaits satisfactorily and can be effective in achieving accurate monitoring and gait assistance for Parkinson’s disease patients during daily life and rehabilitation therapy. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2025.

关键词： Accelerometers

来源：评论

学校读者我要写书评

暂无评论

Energy efficient waste collection vehicle routing system using time series prediction

引用

Multimedia Tools and Applications 2025年第9期84卷 6157-6186页

作者： Kapadia, Neha Mehta, Rupa Computer Engineering Department Sardar Vallabhbhai National Institute of Technology University in Surat Gujarat India IT Department Thakur College of Engineering and Technology Mumbai India

The focus of this research is on addressing the inefficient management of Municipal Solid Waste (MSW), which often results in large quantities of waste being dumped into garbage bins. Previous approaches have struggled to effectively manage MSW promptly. To improve this process, we propose a time-series-based, energy-efficient waste prediction model. Initially, we convert the Bins Historical Time Series Dataset into a Comma-Separated Values (CSV) file format. The model, named Exponential Auto-Regressive Integrated Moving Average-Nesterov-accelerated Adaptive Moment estimation-Long-Short Term Memory (EARIMA-LSTM) using NADAM optimizer, is trained and tested using numeralized features extracted from the dataset. Smart bins equipped with Internet of Things (IoT) technology are then deployed to predict waste levels using the EARIMA-NADAM-LSTM model. After waste level detection, we employ the Scramble Mutation Coral Reefs Optimization-Haversine distance-based K-Means (SMCRO-HK-Means) clustering technique to group the bins. Next, we establish multiple paths between bins nearing capacity using the Ad-hoc On-Demand Distance Vector (AODV) protocol. Finally, the Optimal Path (OP) is determined using the modified A* algorithm based on multi-weight heuristics (Op A*). This approach effectively manages waste collection by preventing bin overflow. Our model achieves good accuracy for waste prediction model with a Mean Absolute Error (MAE) of 0.138%, and 42,518 ms for clustering 500 waste bins, outperforming existing methods. Also, it finds the dynamic optimal route for waste collection with optimal cost. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

VoteDroid: a new ensemble voting classifier for malware detection based on fine-tuned deep learning models

引用

Multimedia Tools and Applications 2025年第12期84卷 10923-10944页

作者： Bakır, Halit Faculty of Engineering and Natural Sciences Department of Computer Engineering Sivas University of Science and Technology Sivas Turkey

In this work, VoteDroid a novel fine-tuned deep learning models-based ensemble voting classifier has been proposed for detecting malicious behavior in Android applications. To this end, we proposed adopting the random search optimization algorithm for deciding the structure of the models used as voter classifiers in the ensemble classifier. We specified the potential components that can be used in each model and left the random search algorithm taking a decision about the structure of the model including the number of each component that should be used and its location in the structure. This optimization method has been used to build three different deep learning models namely CNN-ANN, pure CNN, and pure ANN. After selecting the best structure for each DL model, the selected three models have been trained and tested using the constructed image dataset. Afterward, we suggested hybridizing the fine-tuned three deep-learning models to form one ensemble voting classifier with two different working modes namely MMR (Malware Minority Rule) and LMR (Label Majority Rule). To our knowledge, this is the first time that an ensemble classifier has been fine-tuned and hybridized in this way for malware detection. The results showed that the proposed models were promising, where the classification accuracy exceeded 97% in all experiments. © The Author(s) 2024.

关键词： Android malware

来源：评论

学校读者我要写书评

暂无评论

YOLOCSP-PEST for Crops Pest Localization and Classification

引用

computers, Materials & Continua 2025年第2期82卷 2373-2388页

作者： Farooq Ali Huma Qayyum Kashif Saleem Iftikhar Ahmad Muhammad Javed Iqbal Department of Software Engineering University of Engineering and TechnologyTaxila47050Pakistan Department of Computer Science&Engineering College of Applied Studies&Community ServiceKing Saud UniversityRiyadh11362Saudi Arabia Department of Information Technology Faculty of Computing and Information TechnologyKing Abdulaziz UniversityJeddah21589Saudi Arabia Department of Computer Science University of Engineering and TechnologyTaxila47050Pakistan

Preservation of the crops depends on early and accurate detection of pests on crops as they cause several diseases decreasing crop production and quality. Several deep-learning techniques have been applied to overcome the issue of pest detection on crops. We have developed the YOLOCSP-PEST model for Pest localization and classification. With the Cross Stage Partial Network (CSPNET) backbone, the proposed model is a modified version of You Only Look Once Version 7 (YOLOv7) that is intended primarily for pest localization and classification. Our proposed model gives exceptionally good results under conditions that are very challenging for any other comparable models especially conditions where we have issues with the luminance and the orientation of the images. It helps farmers working out on their crops in distant areas to determine any infestation quickly and accurately on their crops which helps in the quality and quantity of the production yield. The model has been trained and tested on 2 datasets namely the IP102 data set and a local crop data set on both of which it has shown exceptional results. It gave us a mean average precision (mAP) of 88.40% along with a precision of 85.55% and a recall of 84.25% on the IP102 dataset meanwhile giving a mAP of 97.18% on the local data set along with a recall of 94.88% and a precision of 97.50%. These findings demonstrate that the proposed model is very effective in detecting real-life scenarios and can help in the production of crops improving the yield quality and quantity at the same time.

关键词： Deep learning classification of pests YOLOCSP-PEST pest detection

来源：评论

学校读者我要写书评

暂无评论

Improved employee bee algorithm for team formation

引用

Evolving Systems 2025年第2期16卷 1-20页

作者： Shingade, Sandip Niyogi, Rajdeep Shelke, Sushila Department of Computer Science and Engineering Indian Institute of Technology Roorkee Uttarakhand Roorkee247667 India Department of Computer Engineering and Information Technology Veermata Jijabai Technological Institute Matunga Maharashtra Mumbai400019 India Department of Computer Engineering Cummins College of Engineering for Women Karve Nagar Maharashtra Pune411052 India

Finding an appropriate subset of agents (a team) from a larger pool of agents (the source set) so that the team exhibits a desired quality is the essence of the team formation problem. This problem is recognized to have high computational complexity, classified as NP-hard. In order to reduce the state space, this research paper incorporates techniques including vertical and horizontal reductions, which convert data to binary representation. To simplify the data representation for analysis, these methods use vertical and horizontal reductions to convert data to binary format, generating a subgraph. With this, the Honeybee Optimization algorithm’s Employee Bee phase (EBP) swap operation characteristic is used to efficiently optimize team communication costs by making it easier to determine a team with a minimum communication costs (MCC). The proposed approach is assessed using two well-known real-world data sets, the ACM data set and the DBLP data set, demonstrating an improvement of 79% in the ACM dataset and 68% in the DBLP dataset compared to existing literature approaches like Improved Jaya Modified Swap Operation, Improved Grey Wolf Optimization, and State Space Reduction Techniques for Team Formation. This highlights the significant positive impact of state space reduction techniques combined with the IEB on team performance, particularly enhancing communication efficiency and overall team effectiveness, as measured by MCC metrics. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2025.

关键词： NP-hard

来源：评论

学校读者我要写书评

暂无评论

An Integrated Framework with Enhanced Primitives for Post-Quantum Cryptography: HEDT and ECSIDH for Cloud Data Security and Key Exchange

Informatica (Slovenia)

引用

Informatica (Slovenia) 2025年第11期49卷 165-176页

作者： Ilias, Shaik Mohammad Sharmila, V. Ceronmani Durga, V. Sathya Dept. of Computer Science and Engineering Hindustan Institute of Technology and Science Chennai India Department of Information Technology Hindustan Institute of Technology and Science Chennai India Department of Computer Science&Engineering Hindustan Institute of Technology and Science Chennai India

If adversaries were to obtain quantum computers in the future, their massive computing power would likely break existing security schemes. Since security is a continuous process, more substantial security schemes must be developed. Current PQC schemes primarily focus on data security or key exchange, and further improvement towards enhanced PQC primitives is required. Our proposal in this research is an innovative paradigm for PQC-focused cloud data security. The proposed HEDT approach achieves encryption and decryption with significantly lower latency (20% improvement) and higher reliability than AES, DES, and RSA, as demonstrated through experimental results. Furthermore, ECSIDH, a hybrid key exchange mechanism combining SIDH and ECDH, improves security strength by 50% while maintaining computational costs within 1.13x of SIDH. Compared to individual key exchange schemes like SIDH, ECSIDH offers superior security as a PQC candidate. These results confirm the robustness and efficiency of the proposed framework in ensuring secure data outsourcing and key exchange in cloud environments. © 2025 Slovene Society Informatika. All rights reserved.

关键词： Quantum computers

来源：评论

学校读者我要写书评

暂无评论

3-D AAV Deployment for Maximizing Harvested Energy in WRSNs: A Soft Actor-Critic Approach

引用

IEEE Internet of Things Journal 2025年第9期12卷 12083-12093页

作者： Bhalerao, Nilesh Vijaykumar Mothku, Sai Krishna Raju, Mekala Ratna National Institute of Technology Tiruchirappalli Department of Computer Science and Engineering Tiruchirappalli620015 India SRM University AP Department of Computer Science and Engineering Amaravati522240 India

Autonomous aerial vehicles (AAVs) with wireless power transfer (WPT) technology offer a promising solution to extend the lifetime of wireless rechargeable sensor networks (WRSNs) by swiftly recharging multiple sensor nodes. Existing strategies often prioritize optimizing the AAV trajectory and charging schedule. While overlooking diverse energy needs and spatial distribution of sensor nodes, many approaches rely on positioning the AAV at the center of the region above a fixed height. This also leads to potential energy wastage by unnecessarily recharging already adequately powered nodes near the AAV. A key challenge lies in optimizing the 3-D position of the AAV to maximize energy harvested by sensor nodes, considering their diverse energy demands and spatial distribution. To tackle this challenge, we propose a novel approach that jointly optimizes the 3-D deployment of the AAV and energy harvested by sensor nodes. We formulate this complex, nonconvex problem as a Markov decision process (MDP) and leverage deep reinforcement learning based on the soft actor-critic (SAC) algorithm to address it. The AAV learns to navigate continuous state and action spaces, enabling it to discover optimal hovering locations. Our novel reward function allows the AAV to prioritize energy-critical sensor nodes and regions with higher sensor node density, maximizing energy harvesting within limited hovering times. Simulation results demonstrate that our proposed method significantly outperforms conventional AAV deployment strategies, substantially improving sensor node energy levels and network sustainability. © 2014 IEEE.

关键词： Antennas

来源：评论

学校读者我要写书评

暂无评论

A survey on cross-user federated recommendation

引用

Science China(Information Sciences) 2025年第4期68卷 7-32页

作者： Enyue YANG Yudi XIONG Wei YUAN Weike PAN Qiang YANG Zhong MING College of Computer Science and Software Engineering Shenzhen University School of Electrical Engineering and Computer Science The University of Queensland WeBank AI Lab WeBank Department of Computer Science and Engineering Hong Kong University of Science and Technology College of Big Data and Internet Shenzhen Technology University Guangdong Laboratory of Artificial Intelligence and Digital Economy (SZ)

Recommender systems are effective in mitigating information overload, yet the centralized storage of user data raises significant privacy concerns. Cross-user federated recommendation(CUFR) provides a promising distributed paradigm to address these concerns by enabling privacy-preserving recommendations directly on user devices. In this survey, we review and categorize current progress in CUFR, focusing on four key aspects: privacy, security, accuracy, and efficiency. Firstly,we conduct an in-depth privacy analysis, discuss various cases of privacy leakage, and then review recent methods for privacy protection. Secondly, we analyze security concerns and review recent methods for untargeted and targeted *** untargeted attack methods, we categorize them into data poisoning attack methods and parameter poisoning attack methods. For targeted attack methods, we categorize them into user-based methods and item-based methods. Thirdly,we provide an overview of the federated variants of some representative methods, and then review the recent methods for improving accuracy from two categories: data heterogeneity and high-order information. Fourthly, we review recent methods for improving training efficiency from two categories: client sampling and model compression. Finally, we conclude this survey and explore some potential future research topics in CUFR.

关键词： cross-user federated recommendation federated recommendation federated learning recommender systems user privacy

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：