检索结果-内蒙古大学图书馆

Advances in DeepFake detection algorithms: Exploring fusion techniques in single and multi-modal approach

INFORMATION fusion 2025年 118卷

作者： Kumar, Ashish Singh, Divya Jain, Rachna Jain, Deepak Kumar Gan, Chenquan Zhao, Xudong Bennett Univ Sch Comp Sci Engn & Technol Greater Noida Uttar Pradesh India Bhagwan Parshuram Inst Technol New Delhi India Dalian Univ Technol Key Lab Intelligent Control & Optimizat Ind Equipm Minist Educ Dalian 116024 Peoples R China Symbiosis Int Univ Symbiosis Inst Technol Pune 412115 India Chongqing Univ Posts & Telecommun Sch Cyber Secur & Informat Law Chongqing 400065 Peoples R China

In recent years, generative artificial intelligence has gained momentum and created extremely realistic synthetic multimedia content that can spread misinformation and mislead society. Deepfake detection is a technique consisting of frameworks, algorithms and approaches to predict manipulated contents namely, image, audio and video. To this end, we have analyzed and explored various deepfake detection frameworks by categorizing them as single-modal or multi-modal approaches. To provide better understanding and clarity, single-modal approaches are further categorized as conventional and advanced techniques. Conventional techniques extract complementary handcrafted features and classify them using machine-learning-based algorithms. On the other hand, advanced techniques adopt deep learning and hybrid algorithms to detect deepfakes. Multi-modal techniques utilize a mixture of two or more modalities for feature extraction and fuse them to obtain the final classification scores. These techniques are also categorized either as deep learning or hybrid techniques. The complementary features, multiple modalities, and deep learning models are fused adaptively using score-level or feature-level fusion. The advantages, features, practical applications, and limitations under each category are highlighted to address the challenges and determine future trends to counter deepfakes. In addition, recommendations are also elaborated to evaluate the potential of artificial intelligence in deepfake detection for providing a safer and more reliable digital world.

关键词： DeepFake Artificial intelligence Generative adversarial network fusion algorithms Transformer Detection

来源：评论

学校读者我要写书评

暂无评论

SC-MLIDS: fusion-based Machine Learning Framework for Intrusion Detection in Wireless Sensor Networks

引用

AD HOC NETWORKS 2025年 175卷

作者： Zhang, Hongwei Upadhyay, Darshana Zaman, Marzia Jain, Achin Sampalli, Srinivas Dalhousie Univ Fac Comp Sci Halifax NS B3H 4R2 Canada Cistel Technol Inc Res & Dev Dept Ottawa ON K2E 7V7 Canada Norleaf Networks Res & Dev Dept Gatineau PQ J8Y 2V5 Canada

This paper proposes the Server-Client Machine Learning Intrusion Detection System (SC-MLIDS), a novel fusion framework designed to enhance security in Wireless Sensor Networks (WSNs), which are inherently vulnerable to various security threats due to their distributed nature and resource constraints. Traditional Intrusion Detection Systems (IDSs) often face challenges with high computational demands and privacy issues. SC-MLIDS addresses these problems by integrating Federated Learning (FL) with a multi-sensor fusion approach to implementing two layers of defence that operate independently of specific attack types. Moreover, this framework leverages a server-client architecture to efficiently manage and process data from sensor nodes, sink nodes, and gateways within the network. The core innovation of SC-MLIDS lies in its dual model aggregation algorithms at the gateway: one assesses model performance and weight, while the other uses majority voting to integrate predictions from both client and server models. As a result, this approach reduces redundant data transmissions and enhances detection accuracy, making it more effective than conventional methods in WSNs. Our proposed framework outperforms current state-of-the-art techniques, achieving F1-scores of 99.78% and 98.80% for the two aggregation algorithms, namely, Weighted Score and Majority Voting. This validation demonstrates the effectiveness of SC-MLIDS in providing accurate intrusion detection and robust data management.

关键词： Wireless sensor networks Machine learning Intrusion detection systems Ensemble learning Federated learning Server-client models Model aggregation Network security fusion algorithms

来源：评论

学校读者我要写书评

暂无评论

A sensor data fusion algorithm based on suboptimal network powered deep learning

引用

ALEXANDRIA ENGINEERING JOURNAL 2022年第9期61卷 7129-7139页

作者： Pan, Ning Zhengzhou Tourism Coll Sch Informat Engn Zhengzhou 450000 Henan Peoples R China

To improve the performance of data fusion in wireless sensor networks, based on the suboptimal network powered deep learning model, a data fusion algorithm that combines a stacked autoencoder (SAE) and a clustering protocol is proposed. The fusion algorithm for sensor data is investigated by suboptimal network powered deep learning to analyse the fusion processing at the sensor measurement level and the fusion processing at the localization result level. This paper establishes a suboptimal network powered deep learning model and derives its weighted least squares solution algorithm based on this model, also, the best linear unbiased estimation method for fusion of vibration sensor and vision sensor localization results is adopted. First, to overcome the filter divergence caused by rounding error, reduce the computational complexity, ensure the nonnegativity of the covariance, and improve the convergence speed of the filter, the target state volume after fusion using the proposed model is closer to the real state volume of the target compared to the state volume of each local sensor detection, verify the validity of the algorithm was increased. (c) 2021 THE AUTHORS. Published by Elsevier BV on behalf of Faculty of Engineering, Alexandria University This is an open access article under the CC BY-NC-ND license (http://***/

关键词： Suboptimal networks Deep learning Sensor data fusion algorithms

来源：评论

学校读者我要写书评

暂无评论

ACTIVE AND PASSIVE L-BAND MICROWAVE REMOTE SENSING FOR SOIL MOISTURE -A TEST-BED FOR SMAP fusion algorithms

ACTIVE AND PASSIVE L-BAND MICROWAVE REMOTE SENSING FOR SOIL ...

引用

IEEE Joint International Geoscience and Remote Sensing Symposium (IGARSS) / 35th Canadian Symposium on Remote Sensing

作者： Montzka, C. Bogena, H. Jagdhuber, T. Hajnsek, I. Horn, R. Reigber, A. Hasan, S. Ruediger, C. Jaeger, M. Vereecken, H. Res Ctr Julich Inst Bio & Geosci Agrosphere IBG 3 Julich Germany

ISBN: (纸本)9781479957750

The objective of the NASA Soil Moisture Active & Passive (SMAP) mission is to provide global measurements of soil moisture and its freeze/thaw state. The SMAP measurement approach is to integrate L-band radar and radiometer as a single observation system combining the respective strengths of active and passive remote sensing for enhanced soil moisture mapping. Airborne instruments will be a key part of the SMAP validation program. Here, we present an airborne campaign in the Rur catchment, Germany, in which the passive L-band system Polarimetric L-band Multi-beam Radiometer (PLMR2) and the active L-band system DLR F-SAR were flown on six dates in 2013. The flights covered the full heterogeneity of the area under investigation, i.e. all types of land cover and experimental monitoring sites. The obtained data sets are used as a test-bed for the analysis of existing and development of new active-passive fusion techniques.

关键词： SMAP Airborne Microwave Remote Sensing fusion algorithms Soil Moisture

来源：评论

学校读者我要写书评

暂无评论

Algorithm Research Based on 2D LiDAR-Binocular Camera fusion 6

Algorithm Research Based on 2D LiDAR-Binocular Camera Fusion

引用

6th IEEE Advanced Information Technology, Electronic and Automation Control Conference (IEEE IAEAC)

作者： Wang, Chuyue Wang, Meiling Zhan, Wenfeng Ye, Xiaodong Chinese Acad Sci Hefei Inst Phys Sci Hefei Anhui Peoples R China Univ Sci & Technol China Hefei Hefei Anhui Peoples R China

ISBN: (数字)9781665458641

ISBN: (纸本)9781665458641

Autonomous navigation algorithms for mobile robots are widely used. At present, the mainstream civil indoor mobile robot autonomous navigation algorithms are using low cost 2D LiDAR, which obtains less environmental information. In this paper, a lower-cost binocular camera and 2D LiDAR are used as sensors to obtain more environmental information. Firstly, the binocular camera is calibrated and fitted with depth error. Secondly, the 3D information of the binocular camera is intercepted and projected to obtain two-bit point cloud data. Finally, the processed 2D camera point cloud data is fitted with the LiDAR point cloud for interpolation, and the fused 2D point cloud is obtained for autonomous robot navigation. The experimental verification can determine that this algorithm has significantly improved the autonomous navigation effect compared to using 2D LiDAR alone.

关键词： Autonomous Navigation fusion algorithms Lidar Binocular depth camera

来源：评论

学校读者我要写书评

暂无评论

Multimodal recognition of pain intensity and pain modality with machine learning

引用

SCHMERZ 2020年第5期34卷 400-409页

作者： Walter, S. Al-Hamadi, A. Gruss, S. Frisch, S. Traue, H. C. Werner, P. Univ Klinikum Ulm Klin Psychosomat Med & Psychotherapie Sekt Med Psychol Frauensteige 6 D-89075 Ulm Germany Otto von Guericke Univ Inst Informat & Kommunikat Tech Fachgebiet Neuroinformat Tech Magdeburg Germany Praxis Neurol & Psychiat Leutkirch Germany

Background The objective recording of subjectively experienced pain is a problem that has not been sufficiently solved to date. In recent years, data sets have been created to train artificial intelligence algorithms to recognize patterns of pain intensity. The multimodal recognition of pain with machine learning could provide a way to reduce an over- or undersupply of analgesics, explicitly in patients with limited communication skills. Objectives This study investigated the methodology of automated multimodal recognition of pain intensity and modality using machine-learning techniques of artificial intelligence. Multimodal recognition rates of experimentally induced phasic electrical and heat pain stimuli were compared with uni- and bimodal recognition rates. Material and methods On the basis of the X-ITE Pain Database, healthy subjects were stimulated with phasic electro-induced pain and heat pain, and their corresponding pain responses were recorded with multimodal sensors (acoustic, video-based, physiological). After complex signal processing, machine-learning methods were used to calculate recognition rates with respect to pain intensity (baseline vs. pain threshold, pain tolerance, mean value of pain threshold and tolerance) and pain modality (electrical vs. heat). Finally, a statistical comparison of uni- vs. multimodal and bi- vs. multimodal detection rates was performed. Results With few exceptions, multimodal recognition of pain intensity rates was statistically superior to unimodal recognition rates, regardless of the pain modality. Multimodal pain recognition distinguished significantly better between heat and electro-induced pain. Further, multimodal recognition rates were predominantly superior to bimodal recognition rates. Conclusion Priority should be given to the multimodal approach to the recognition of pain intensity and modality compared with unimodality. Further clinical studies should clarify whether multimodal automated recognition of pain in

关键词： Automated pain recognition Machine learning Artificial intelligence Multimodality fusion algorithms

来源：评论

学校读者我要写书评

暂无评论

Multiresolution Compressive Feature fusion for Spectral Image Classification

引用

IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING 2019年第12期57卷 9900-9911页

作者： Marcos Ramirez, Juan Arguello, Henry Univ Los Andes Dept Elect Engn Merida 5101 Venezuela Univ Ind Santander Dept Comp Sci Bucaramanga 680002 Colombia

Compressive spectral imaging (CSI) has emerged as an alternative acquisition framework that simultaneously senses and compresses spectral images. In this context, the spectral image classification from CSI compressive measurements has become a challenging task since the feature extraction stage usually requires reconstructing the spectral image. Moreover, most approaches do not consider multi-sensor compressive measurements. In this paper, an approach for fusing features obtained from multi-sensor compressive measurements is proposed for spectral image classification. To this end, linear models describing low-resolution features as degraded versions of the high-resolution features are developed. Furthermore, an inverse problem is formulated aiming at estimating high-resolution features including both a sparsity-inducing term and a total variation (TV) regularization term to exploit the correlation between neighboring pixels of the spectral image, and therefore, to improve the performance of pixel-based classifiers. An algorithm based on the alternating direction method of multipliers (ADMM) is described for solving the fusion problem. The proposed feature fusion approach is tested for two CSI architectures: three-dimensional coded aperture snapshot spectral imaging (3D-CASSI) and colored CASSI (C-CASSI). Extensive simulations on various spectral image data sets show that the proposed approach outperforms other classification approaches under different performance criteria.

关键词： Feature extraction Image coding Apertures Sensors Optical imaging Image resolution Compressive spectral imaging (CSI) feature fusion fusion algorithms multi-sensor spectral image classification

来源：评论

学校读者我要写书评

暂无评论

Evaluation of commercial available fusion algorithms for Geoeye Data

Evaluation of commercial available fusion algorithms for Geo...

引用

Conference on Earth Resources and Environmental Remote Sensing/GIS Applications IV

作者： Vaiopoulos, Aristides D. Nikolakopoulos, Konstantinos G. Univ Athens Dept Geol GR-10679 Athens Greece

ISBN: (纸本)9780819497635

In this study ten commercial available fusion techniques and more especially the Ehlers, Gram-Schmidt, High Pass Filter, Local Mean Matching (LMM), Local Mean and Variance Matching (LMVM), Modified IHS (ModIHS), Pansharp, PCA, HCS (Hyperspherical Color Space) and Wavelet were used for the fusion of Geoeye panchromatic and multispectral data. The panchromatic data have a spatial resolution of 0.5m while the multispectral data have a spatial resolution of 2.0m. The optical result, the statistical parameters and different quality indexes such as ERGAS, Q, entropy were examined and the results are presented. The broader area of Pendeli mountain near to the city of Athens Greece and more especially two sub areas with different characteristics were chosen for the comparison. The first sub area is located at the edge of the urban fabric and combines at the same time the characteristics of an urban and a rural area. The second sub area comprises a large open quarry and it is suitable to examine which fused product is more suitable for mine monitoring.

关键词： Geoeye fusion algorithms quality control

来源：评论

学校读者我要写书评

暂无评论

Obstacle information detection method based on multiframe three-dimensional lidar point cloud fusion

引用

OPTICAL ENGINEERING 2019年第11期58卷 116102-116102页

作者： Li, Jing Li, Rui Wang, Junzheng Yan, Min Beijing Inst Technol Sch Automat Key Lab Complex Syst Intelligent Control & Decis Beijing Peoples R China Beijing Inst Technol Sch Automat Beijing Peoples R China

An obstacle detection method based on multiframe point cloud fusion and ground plane estimation is proposed. First, the robot, three-dimensional lidar, and two-axis rotating platform motion models were established. With the pitch rotation of lidar and the movement of the robot, the multiframe point cloud data were superposed to obtain a dense point cloud. Then, the incremental linear fitting method was adopted to estimate the ground plane to eliminate the ground disturbance. Finally, the pairwise linkage clustering method could cluster multiple obstacles and the projection method could obtain the specific size information of obstacles. Therefore, verification of experiments was carried on the Robotics Lab Platform (BIT-NAZA), experimental results showed that multiframe point cloud data can be accurately integrated to obtain dense point cloud and then get information of obstacles. For some low-height obstacles with sparse density of point clouds, the detection accuracy of an object with a height of about 10 cm is about 95%, whereas for other higher objects, the detection rate will be higher. So the proposed algorithm can achieve its detection and related parameter acquisition efficiently. (C) 2019 Society of Photo-Optical Instrumentation Engineers (SPIE)

关键词： obstacle detection fusion algorithms ground plane estimation lidar mobile robots

来源：评论

学校读者我要写书评

暂无评论

Functional clusters analysis and research based on differential coexpression networks

引用

BIOTECHNOLOGY & BIOTECHNOLOGICAL EQUIPMENT 2018年第1期32卷 171-182页

作者： Meng, Shuai Liu, Guixia Su, Lingtao Sun, Liyan Wu, Di Wang, Lingwei Zheng, Zhao Jilin Univ Coll Comp Sci & Technol Changchun Jilin Peoples R China Jilin Univ Minist Educ Key Lab Symbol Computat & Knowledge Engn Changchun Jilin Peoples R China

Differential coexpression analysis has gradually become an important approach to improve the conventional method of analysing differentially expressed genes. With this approach, it is possible to discover disease mechanisms and underlying regulatory dynamics which remain obscure in differential expression analysis. The detection of differential coexpression links and functional clusters between different disease states is a demanding task. Nevertheless, there is no gold standard for detecting differential coexpression links and functional clusters. Consequently, we developed a novel fusion algorithm FDvDe (fusion of differential vertex and differential edge) to detect differential coexpression links by aggregating the set of 'differential vertex' and 'differential edge.' Then, we constructed differential coexpression networks between normal and tumour states by integrating the differential coexpression links. With this approach, we identified 1823 genes and 29370 links. Then, we developed the algorithms GTHC (GO term hierarchical clusters) to identify functional modules. The distance matrix used in the hierarchical process was formed by the GO semantic similarity. Furthermore, we aggregated the densities among clusters describing the connectivity among clusters and topological property analysis to discover the hub genes and hub pathways which play an important role in disease mechanism. In this paper, we showed that our approach worked well on a data set of breast cancer samples (68 tumour samples) and normal samples (73 normal samples), and revealed the crucial role and biological significance of the modules and hub genes found in this approach.

关键词： Differential coexpression network fusion algorithms GO hierarchical clusters topological property analysis hub gene

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：