检索结果-内蒙古大学图书馆

Computational Intelligence, Communication Technology and Networking (CICTN), International Conference on

作者： Pradum Kumar Singh Sourav Sharma Ritu Sachdeva School of Computer Applications Manav Rachna International Institute of Research & Studies Faridabad India

ISBN: (数字)9798331530389

ISBN: (纸本)9798331530396

The proliferation of deepfakes-also termed artificial intelligence-generated synthetic media poses unprecedented challenges to the digital authenticity of media, for media integrity, and to societies’ trust in its credibility. Despite such significant technological advancements, current methodologies for detecting deepfakes remain somewhat fragmented, reactive, and often unable to keep pace with rapidly evolving technologies in generative AI. This article considers a comprehensive multi-modal approach toward deepfake detection by integrating an advanced machine learning algorithm, novel statistical correlation techniques, and forensic image analysis that would fill in the critical gaps in the existing frameworks. By using an extensive dataset of 10,000 synthetic and authentic media samples from diverse domains, we developed a hybrid neural network architecture that attained 94.3% accuracy on identifying AI-generated content. Our methodology utilized ensemble learning by combining spatial-temporal inconsistency detection, biological signal analysis, and advanced feature extraction algorithms. Key findings demonstrate significant vulnerabilities in current deepfake generation models, with our proposed technique successfully identifying subtle artifacts and inconsistencies across multiple generative AI platforms. This research conclusively demonstrates that proactive, adaptive detection strategies are critical to mitigating the potential risks associated with synthetic media, providing a robust framework for future technological interventions in digital forensics and media authentication.

关键词： Deepfakes Machine learning algorithms Generative AI neural networks Digital forensics Signal processing algorithms Authentication Media Feature extraction Information integrity

来源：评论

学校读者我要写书评

暂无评论

applications of artificial neural networks in image processing IV : 28-29 January 1999, San Jose, California / Nasser M. Nasrabadi, Aggelos K. Katsaggelos, chairs/editors;sponsored by IS&T--the Soci

Applications of artificial neural networks in image processi...

引用

1999年

作者： Nasser M. Nasrabadi Aggelos K. Katsaggelos chairs/editors sponsored by IS&T--the Society for Imaging Science and Technology [and] SPIE--the International Society for Optical Engineering.

来源：评论

学校读者我要写书评

暂无评论

Optimizing CNN Architecture for Quality Control of Corneal Confocal Microscopy images Using a Genetic Algorithm 15th

Optimizing CNN Architecture for Quality Control of Corneal C...

引用

15th International Conference on Soft Computing and Pattern Recognition, SoCPaR 2023 and 14th World Congress on Nature and Biologically Inspired Computing, NaBIC 2023

作者： Bönhof, Berenice N. Hanne, Thomas Dornberger, Rolf Gachnang, Phillip Bönhof, Gidon J. School of Business University of Applied Sciences and Arts Northwestern Switzerland Basel Switzerland Vetsuisse-Faculty Institute of Veterinary Pharmacology and Toxicology University of Zurich Zurich Switzerland Institute for Information Systems University of Applied Sciences and Arts Northwestern Switzerland Olten Switzerland Institute for Clinical Diabetology German Diabetes Center Leibniz Center for Diabetes Research at Heinrich Heine University Düsseldorf Düsseldorf Germany Department of Endocrinology and Diabetology Medical Faculty University Hospital Düsseldorf Heinrich Heine University Düsseldorf Düsseldorf Germany German Center for Diabetes Research Partner Düsseldorf München-Neuherberg Germany

ISBN: (纸本)9783031810824

The use of artificial Intelligence (AI) in the medical field has grown significantly in recent years, also with computer vision techniques, such as Convolutional neural networks (CNNs), gaining prominence in medical image processing applications. While CNNs have proven helpful in various medical tasks, there is a research gap in the automatic selection of Corneal Confocal Microscopy (CCM) images for diagnostic purposes. CCM could be used to screen for early stages of Diabetic Peripheral Neuropathy (DPN), the most common complication of diabetes. However, manually selecting CCM images for diagnosis is time-consuming and requires expertise. This paper aims to evaluate the effectiveness of a Genetic Algorithm (GA) in finding an optimized CNN architecture for accurate quality control of CCM images. The GA-based approach is compared to the widely used VGG16 architecture. The results demonstrate that the GA-evolved CNN architecture achieved a validation accuracy of 67.57 to 72.97%, outperforming VGG16. This research highlights the potential of GAs in optimizing CNN architectures for medical image classification tasks and indicates a first step towards automatically selecting CCM images for analysis. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Convolutional neural Network Corneal Confocal Microscopy Diabetic Peripheral Neuropathy Genetic Algorithm VGG16

来源：评论

学校读者我要写书评

暂无评论

Generalized fractional optimization-based explainable lightweight CNN model for malaria disease classification

引用

Computers in Biology and Medicine 2025年 185卷 109593-109593页

作者： Khan, Zeshan Aslam Waqar, Muhammad Raja, Muhammad Junaid Ali Asif Chaudhary, Naveed Ishtiaq Khan, Abeer Tahir Mehmood Anwar Raja, Muhammad Asif Zahoor International Graduate Institute of Artificial Intelligence National Yunlin University of Science and Technology Section 3 123 University Road Douliou Yunlin64002 Taiwan Department of Computer Science and Information Engineering National Yunlin University of Science and Technology Section 3 123 University Road Douliou Yunlin64002 Taiwan Future Technology Research Center National Yunlin University of Science and Technology Section 3 123 University Road Douliou Yunlin64002 Taiwan Rawal Institute of Health Sciences Islamabad Pakistan

Over the past few decades, machine learning and deep learning (DL) have incredibly influenced a broader range of scientific disciplines. DL-based strategies have displayed superior performance in image processing compared to conventional standard methods, especially in healthcare settings. Among the biggest threats to global public health is the fast spread of malaria. The plasmodium falciparum infection, the disease origin causes the intestinal illness. Fortunately, advances in artificial intelligence techniques have made it possible to use visual data sets to quickly and effectively diagnose malaria which has also proven to be cost and time effective. In literature, several DL approaches have previously been used with good precision but suffer from computational inefficiency and interpretability. Therefore, this research proposes a generalized fractional order-based explainable lightweight convolutional neural network model to overcome these limitations. The fractional order optimization algorithms have proven worth in terms of estimation accuracy and convergence speed for different applications. The proposed fractional order optimizer-based model offers an improved solution to malaria disease diagnosis with a percentage accuracy of 95 % using the standard NIH dataset and outperforms the existing complex models concerning speed and effectiveness. The proposed fractionally optimized lightweight CNN model has shown substantial performance on the external MP-IDB dataset and M5 test set as well by achieving a generalized test accuracy of 92 % and 90.4 % which verifies the robustness and generalizability of the proposed solution under available circumstances. Moreover, the efficacy of the proposed lightweight architecture is endorsed through evaluation metrics of precision, recall, and F1-score. © 2024 Elsevier Ltd

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

applications of artificial neural networks in image processing III : proceedings, 26-27 January 1998, San Jose, California / Nasser M. Nasrabadi & Aggelos K. Katsaggelos, Chairs/Editors;Sponsored by

Applications of artificial neural networks in image processi...

引用

1998年

作者： Nasser M. Nasrabadi & Aggelos K. Katsaggelos Chairs/Editors Sponsored by IS&T-The Society for Imaging Science and Technology SPIE-The international society for optical engineering.

来源：评论

学校读者我要写书评

暂无评论

多样性负实例生成的跨域人脸伪造检测

引用

中国图象图形学报 2025年第2期30卷 421-434页

作者：张晶许盼刘文君郭晓萱孙芳辽宁师范大学计算机与人工智能学院大连116081

目的深度伪造检测(deepfake detection)通过训练复杂深度神经网络,挖掘更具辨别性的人脸图像表示,获得高精度的检测结果,其是一项确保人脸信息真实、可靠和安全的重要技术。然而,目前流行的模型存在过度依赖训练数据,使模型仅在相同域... 详细信息

目的深度伪造检测(deepfake detection)通过训练复杂深度神经网络,挖掘更具辨别性的人脸图像表示,获得高精度的检测结果,其是一项确保人脸信息真实、可靠和安全的重要技术。然而,目前流行的模型存在过度依赖训练数据,使模型仅在相同域内表现出令人满意的检测性能,在跨领域场景中表现出较低泛化性,甚至使模型失效。因此,如何在有限的训练数据下实现跨域环境中的高效伪造人脸检测,成为亟待解决的问题。基于此,本文提出多样性负实例生成的跨域人脸伪造检测模型(negative instance generation-FFD,NIG-FFD)。方法首先,通过构建孪生自编码网络,获得标签一致的潜在多视图融合特征,引入对比约束提高难样本特征可判别性;其次,在高效训练的同时利用构造规则生成更具多样性的负实例融合特征,提高模型泛化性;最后,构建自适应重要性权值矩阵,避免因负实例生成导致类别分布不平衡使正类别样本欠学习。结果在两个流行的跨域数据集上验证本文模型的有效性,与其他先进方法相比,AUC(area under the receiver operating characteristic curve)值提升了10%。同时,在本域检测中ACC(accuracy score,)与AUC值相比其他方法均提升了近10%与5%。结论与对比方法相比,本文方法在跨域和本域的人脸伪造检测上都取得了优越的性能。本文所提的模型代码已开源至:https://***/LNNU-computer-research-526/NIG-FFD。

关键词：深度伪造检测跨域人脸伪造检测多视图特征融合特征生成对比约束

来源：评论

学校读者我要写书评

暂无评论

人工智能模型栈在国家审计中的应用

引用

审计研究 2025年第1期 51-56页

作者：朱琢江苏省无锡市审计局 214000

近年来,人工智能技术和应用快速发展,ChatGPT等应用不断突破。经历了传统财务审计、计算机辅助审计到大数据审计的审计技术跃迁,无锡市审计局在养老、医保等民生审计项目中,不断寻求人工智能与审计的交汇点。通过多维分析审计对于人工... 详细信息

近年来,人工智能技术和应用快速发展,ChatGPT等应用不断突破。经历了传统财务审计、计算机辅助审计到大数据审计的审计技术跃迁,无锡市审计局在养老、医保等民生审计项目中,不断寻求人工智能与审计的交汇点。通过多维分析审计对于人工智能的功能需求,在审计重点研判、疑点数据分析、文字图像处理等方面探索人工智能技术和工具在项目实施中的应用,提出了人工智能应用功能模型栈理念。不同于常见的一体化人工智能应用产品,提出依托分层低耦合架构,让各类人工智能技术应用融合审计项目全流程,根据项目需求灵活选用生成式人工智能、自然语言处理、机器学习、深度神经网络等技术的审计实践,有效提升审计数据分析效率,减轻了审计人员负担。

关键词：人工智能应用功能模型栈民生审计

来源：评论

学校读者我要写书评

暂无评论

applications of artificial neural networks in image processing

Applications of Artificial Neural Networks in Image Processi...

引用

2001年

作者： Nasser M Nasrabadi Aggelos Konstantinos Katsaggelos

来源：评论

学校读者我要写书评

暂无评论

Improving Student Stress Analysis: Novel Methods for Collecting, Preprocessing, and Fusing Features 3rd

Improving Student Stress Analysis: Novel Methods for Collect...

引用

3rd International Conference on Advancements in Smart Computing and Information Security, ASCIS 2024

作者： Bhatt, Zankhana Dobariya, Ashwin Faculty of Computer Application Marwadi University Gujarat Rajkot India

ISBN: (纸本)9783031862984

Due to the adverse effects of stress on students’ health and academic success, it is to create an effective method for stress detection. This study is used to predict student stress by combining biosignal data with image processing techniques. Employs various image pre processing of facial images in this research like Gaussian Blur, Median Blur, Bilateral Filters, and Canny edge detection to enhance the extraction of features. The Multi-task Cascaded Convolutional neural Network (MTCNN) detects and aligns facia landmarks for precise data collection. Wearable device data on respiration, oxygen, temperature, and heart rate is combined with previously processed images. The robust dataset from physiological and visual characteristics may benefit future stress prediction machine learning applications. This comprehensive real-time stress detection method allows the development of customized interventions to improve student health and academic performance. The author is developing new artificial minority oversampling methods and improving machine learning models. The list includes KNN, Decision Trees, Random Forests, and XG Boost. The author has achieved accuracy rates and F1-macro score over 90% with KNN and SVM and expects even better results with other supervised machine learning algorithms. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Evaluation of Efficiency in Segmenting Barrett's Esophagus images by neural Network Models. Integration in Low-Power Embedded Systems

SSRN

引用

SSRN 2025年

作者： Soriano-Garcia, Miguel S. Sevilla-Escoboza, J. Ricardo Galván-Tejada, Carlos Eric Chiu Coutino, Roger Villagrana Bañuelos, Karen E. Chiu, Roger Center for Research in Mathematics CIMAT AC Campus Zacatecas Calle Lasec y Andador Galileo Galilei Manzana 3 Lote 7 Quantum Ciudad del Conocimiento Zacatecas Zacatecas98160 Mexico Lab. de Metrologia e Instrumentacion CULagos UDG Av. Enrique Diaz de Leon No. 1144 Jalisco Lagos de Moreno47460 Mexico Unidad Académica de Ingeniería Eléctrica Universidad Autónoma de Zacatecas Jardín Juarez 147 Centro Zacatecas Zacatecas Zacatecas98000 Mexico Universidad LaSalle Av. Universidad 602 Guanajuato Leon37150 Mexico Departamento de Ciencias Exactas y Tecnología Centro Universitario de los Lagos Av. Enrique Diaz de Leon No. 1144 Jalisco Lagos de Moreno47460 Mexico Universidad de Guadalajara Mexico

Barrett’s esophagus, caused by prolonged gastroesophageal reflux, raises the risk of esophageal adenocarcinoma. Early detection is vital for better survival rates. Recent GPU advancements enable AI, particularly CNN-based neural networks, to assist in diagnosis. This study evaluates neural models for segmenting Barrett’s lesions in endoscopic images to improve early detection. Model efficiency is assessed using U-Net with various encoders, selecting the best encoder and applying different pre-training weights. Models are compared based on these criteria. The PSPNet model with a ResNet50 encoder pre-trained on imageNet achieves the best validation scores for detecting and segmenting Barrett’s esophagus lesions. When healthy esophageal endoscopic images are introduced, this model outperforms others in detecting the absence of lesions. In conclusion, various models can perform this task, but different encoders and pretraining significantly affect performance. The classic U-Net excels in training but is surpassed by models like PSPNet and PAN in validation. This highlights the importance of choosing appropriate encoders and pretraining strategies to enhance the effectiveness of AI models in diagnosing Barrett’s esophagus lesions for improved patient outcomes. The PSPNet model has also been optimized for execution in low-power embedded systems, such as the Jetson board. With its real-time video processing capabilities, this device leverages the advantages of PSPNet in artificial vision applications. Thanks to this optimization, it is possible to implement efficient and high-performance solutions in resource-constrained environments, maintaining the precision and speed necessary for real-time processing. © 2025, The Authors. All rights reserved.

关键词： image segmentation

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：