检索结果-内蒙古大学图书馆

2024 IEEE International Conference on Modeling, Simulation and Intelligent Computing, MoSICom 2024

作者： Agrawal, Ansh Kaur, Kirandeep Kaur, Harkeerat Indian Institute of Technology Computer Science Engineering Jammu & Kashmir Jammu India Indian Institute of Technology Faculty Computer Science Engineering Jammu & Kashmir Jammu India

ISBN: (纸本)9798331533311

In recent years, the integration of Explainable Artificial Intelligence (XAI) in facial biometrics has gained momentum, driven by a growing need for transparency in automated decision-making systems. This paper explores the application of XAI in face detection and identification to bridge the gap between model accuracy and interpretability, aiming to address challenges unique to biometric systems, including bias mitigation, user trust, and ethical implications. Unlike traditional 'black-box' models, the proposed framework incorporates visual and feature-based explanations that highlight key facial regions contributing to identification outcomes, enabling more transparent model assessments. We evaluate the effectiveness of state-of-the-art XAI techniques, such as SHAP, LIME, and Grad-CAM, through rigorous testing on diverse datasets under varying conditions, including occlusions, lighting, and facial expressions. Additionally, this research introduces a novel hybrid interpretability metric tailored to biometric applications, assessing the quality of explanations based on fidelity, usability, and impact on user trust. Results demonstrate that the proposed XAI-enhanced model offers insights into decision processes without compromising accuracy, advancing the state of explainability in biometric systems. This work underscores the importance of explainability as a means to achieve ethical, reliable, and legally compliant AI systems in sensitive domains like facial biometrics. © 2024 IEEE.

关键词： Biometrics

来源：评论

学校读者我要写书评

暂无评论

KGFakeNet: A Knowledge Graph-Enhanced Model for Fake News Detection 31

KGFakeNet: A Knowledge Graph-Enhanced Model for Fake News De...

引用

2025 Workshop on Generative AI and Knowledge Graphs, GenAIK 2025, co-located with the 31st International Conference on Computational Linguistics, COLING 2025

作者： Kumar, Anuj Kumar, Pardeep Yadav, Abhishek Ahlawat, Satyadev Prasad, Yamuna Department of Computer Science & Engineering Indian Institute of Technology Jammu India Department of Electrical Engineering Indian Institute of Technology Jammu India

ISBN: (纸本)9798891762121

The proliferation of fake news on social media has intensified the spread of misinformation, promoting societal biases, hate, and violence. While recent advancements in Generative AI (GenAI), particularly large language models (LLMs), have shown promise, these models often need more structured representation for accurate verification, as they rely on pre-trained data patterns without access to real-time or validated information. This study presents a framework that utilizes Open Information Extractor 6 (OpenIE6) to extract triplet relationships (subject-predicate-object) from statements and justifications to compute the cosine similarity between the Knowledge Graphs (KGs) of the statements and their supporting justification to precisely measure the relevance and alignment between them. This similarity feature is integrated with an attention mechanism over GenAI-generated embeddings to enhance the model’s ability to capture semantic features accurately. In addition, a Multi-Layer Perceptron (MLP) classifier is employed to integrate all features, resulting in a 4% improvement in accuracy and a 5% increase in F1-score over state-of-the-art LLM-based approaches. ©2025 International Committee on Computational Linguistics (ICCL)

关键词： Knowledge graph

来源：评论

学校读者我要写书评

暂无评论

Emotion Privacy: A Privacy-Preserving Framework for Emotionally Enriched Video

Emotion Privacy: A Privacy-Preserving Framework for Emotiona...

引用

2024 IEEE International Conference on Modeling, Simulation and Intelligent Computing, MoSICom 2024

作者： Kaur, Kirandeep Agrawal, Ansh Kaur, Harkeerat Computer Science Engineering Indian Institute of Technology Jammu & Kashmir Jammu India

ISBN: (纸本)9798331533311

As the use of video systems continues to expand across various domains, addressing privacy concerns while maintaining data integrity has become increasingly challenging. This paper introduces Emotion Privacy, a versatile framework that anonymizes faces in video feeds by replacing them with their own neutral faces. Unlike traditional anonymization techniques that primarily obscure identity but often fail to hide emotional data, our approach ensures compliance with privacy regulations by neutralizing emotional expressions. By applying neutral 3D face models in real time, we achieve complete emotional concealment, thus protecting emotional cues. Experimental validation reveals that Emotion Privacy achieves a great success in anonymization, ensuring that no emotional content can be inferred from the modified videos. This work presents a scalable, privacy-preserving solution that meets contemporary regulatory standards, offering broad applicability across diverse sectors where video content and privacy concerns intersect. © 2024 IEEE.

关键词： Anonymity

来源：评论

学校读者我要写书评

暂无评论

Genetic algorithm based data controlling method using IoT enabled WSNs

引用

Soft Computing 2025年第5期29卷 2465-2482页

作者： Singh, Samayveer Nandan, Aridaman Singh Sikka, Geeta Malik, Aruna Singh, Pradeep Kumar Department of Computer Science and Engineering National Institute of Technology Punjab Jalandhar India Department of Computer Science and Engineering National Institute of Technology Delhi India Department of Computer Science and Engineering Central University of Jammu J&K Jammu India

Internet of Things (IoT) enabled Wireless Sensor Networks (WSNs) is not only constitute an encouraging research domain but also represent a promising industrial trend that permits the development of various IoT-based applications. These applications span a wide range from industry to education, and from military to agriculture. The IoT device plays a significant role in various IoT-based networks, and the functioning of such network depends upon the battery power. Once the devices are deployed in the hostile environments, replacing batteries becomes impractical. Despite a plethora of research addressing this challenge, IoT networks still face issues. In this paper, a genetic algorithm based data monitoring and controlling method using IoT enabled WSNs is proposed by using movable sinks in IoT enabled HWSNs (OptiGeA). The OptiGeA protocol is designed for the election of cluster heads (CHs) by incorporating factors such as density, distance, energy and heterogeneous node capacity into its fitness function. The investigation of OptiGeA is conducted with single sink, multiple static sinks and multiple movable sinks provide an unbiased comparative assessment. The novel deployment technique and multiple mobile sinks approaches are proposed to reduce the transmission distance between the sink and CH during system operation and address hotspot issue. It is evident that the OptiGeA protocol shows an increment of 10.44% compared to the GAOC, whereas with the inclusion of DDC process the OptiGeA-DDC protocol demonstrates a remarkable increase of 48.33% compared to MS-GAOC. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2025.

关键词： Genetic algorithms

来源：评论

学校读者我要写书评

暂无评论

IVIDNet: Intelligent iris vitality detection via weighted prediction score level fusion

引用

Multimedia Tools and Applications 2023年第29期82卷 45959-45981页

作者： Verma, Palak Selwal, Arvind Sharma, Deepika Department of Computer Science and Engineering Indian Institute of Technology Jammu PO-Nagrota Jammu181221 India Department of Computer Science and Information Technology Central University of Jammu Samba181143 India

Iris recognition is one of the most widely used human authentication mechanism that has gained huge popularity due to its security and efficiency. However, recently attackers have breached the security of these systems via synthetically generated fake iris traits using a variety of artefacts such as paper printouts, video attacks, and cosmetic lenses. To countermeasure these attacks, an anti-deception sub-system is instilled between sensor and feature extractor modules of the system. These sub-systems intelligently measure the vitality of the presented iris traits and are popularly known as iris vitality detection (IVID) or iris anti-spoofing techniques. Recently, the deep learning-based models are widely deployed for iris anti-spoofing approaches, but these methods result in additional training overhead as well as require larger training dataset. In this paper, we expound a novel iris vitality detection network (IVIDNet) that employs the robust features of two pre-trained deep convolutional neural networks (DCNN). The extracted features are then integrated via a new weighted score-level fusion technique. To demonstrates the efficacy of the proposed technique, experimental analysis is carried out on benchmark iris liveness detection Notre Dame 2017 and Notre Dame cosmetic lenses dataset (NDCLD) 2015. The proposed IVIDNet exhibits excellent performance in the known attack scenario with an average classification error rate (ACER) of 0.60%. Moreover, IVIDNet shows superior performance in the unseen attacks scenario (i.e. cross-database) with an ACER of 9.72%. In overall, the proposed model outperforms other similar iris anti-spoofing approaches and efficient in terms of computational overhead. © 2023, The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature.

关键词： Biometrics

来源：评论

学校读者我要写书评

暂无评论

A Fast Access Control Method in IoT Using XGB

引用

SN computer science 2024年第8期5卷 1-14页

作者： Tyagi, Surendra Prasad, Yamuna Jinwala, Devesh C. Bhattacharjee, Subhasis Computer Science and Engineering Indian Institute of Technology Jammu Jammu & Kashmir Jagti Jammu 181221 India Computer Science and Engineering Sardar Vallabhbhai National Institute of Technology Ichchhanath Gujarat Surat 395007 India

Application of Machine Learning (ML) has become pivotal for the creation and continuous refinement of access control policies. While the emergence of Deep Learning (DL) based access control is notable, its effectiveness is contingent on selecting an appropriate model tailored to the dataset. Access control datasets mostly comprise tabular data in numeric or text formats with less number of features. Conventional DL methods may not yield high accuracy, especially with limited data availability. We thereby explored and compared various well-established ML methods, including Random Forest, Decision Tree, K-Nearest Neighbors, Extreme Gradient Boosting (XGB), and DL models such as ResNet and Multi-Layer Perception in the context of access control. We propose to use MinMax Scaler transformation instead one hot encoding that resulted in huge saving on memory footprint and training time. We also propose an XGB based access control mechanism that leverages the power of XGBoost patterns which can nicely manage tabular data, missing data, categorical features, and scaling. Moreover, we also propose to use the Euclidean distance to analyse access control query against the parameters of historical data set. This process automatically maps the query into one of the trained XGB models. These novel techniques improved accuracy, reduced training time and memory footprint. Extensive evaluations on both real-world and synthetic datasets demonstrate superiority of the proposed model when compared against state-of-the-art DL models. The training time has been reduced from three hours in DL method to few seconds in our proposed method with significant reduction in memory. © The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd. 2024.

关键词： Access control Deep learning Machine learning ResNet XGB

来源：评论

学校读者我要写书评

暂无评论

Real-time earthquake magnitude prediction using designed machine learning ensemble trained on real and CTGAN generated synthetic data

引用

Geodesy and Geodynamics 2025年第3期16卷 350-368页

作者： Joshi, Anushka Raman, Balasubramanian Mohan, C. Krishna Department of Computer Science and Engineering Indian Institute of Technology Uttarakhand Roorkee India Department of Computer Science and Engineering Indian Institute of Technology Telangana Hyderabad India

The earthquake early warning (EEW) system provides advance notice of potentially damaging ground shaking. In EEW, early estimation of magnitude is crucial for timely rescue operations. A set of thirty-four features is extracted using the primary wave earthquake precursor signal and site-specific information. In Japan's earthquake magnitude dataset, there is a chance of a high imbalance concerning the earthquakes above strong impact. This imbalance causes a high prediction error while training advanced machine learning or deep learning models. In this work, Conditional Tabular Generative Adversarial Networks (CTGAN), a deep machine learning tool, is utilized to learn the characteristics of the first arrival of earthquake P-waves and generate a synthetic dataset based on this information. The result obtained using actual and mixed (synthetic and actual) datasets will be used for training the stacked ensemble magnitude prediction model, MagPred, designed specifically for this study. There are 13295, 3989, and 1710 records designated for training, testing, and validation. The mean absolute error of the test dataset for single station magnitude detection using early three, four, and five seconds of P wave are 0.41, 0.40, and 0.38 MJMA. The study demonstrates that the Generative Adversarial Networks (GANs) can provide a good result for single-station magnitude prediction. The study can be effective where less seismic data is available. The study shows that the machine learning method yields better magnitude detection results compared with the several regression models. The multi-station magnitude prediction study has been conducted on prominent Osaka, Off Fukushima, and Kumamoto earthquakes. Furthermore, to validate the performance of the model, an inter-region study has been performed on the earthquakes of the India or Nepal region. The study demonstrates that GANs can discover effective magnitude estimation compared with non-GAN-based methods. This has a high potential

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Efficient IRS-Aided Rate Optimizations for Dual Functional Radar and Communications

引用

IEEE Communications Letters 2024年第1期28卷 39-42页

作者： Sweta, S. Dubey, Ankit Nguyen, Van-Linh Indian Institute of Technology Jammu Department of Electrical Engineering Jammu and Kashmir Jammu181221 India National Chung Cheng University Department of Computer Science and Information Engineering Chiayi62102 Taiwan

This letter presents a novel method to enable beam steering optimization for secure communications and resilience against radio tracking in an intelligent reflecting surfaces-enabled multiple-input multiple-output dual-functional radar and communications system. The approach involves optimizing beam steering by solving two convex sub-problems: i) maximizing secrecy rate and ii) maximizing legitimate receiver rate, solved using Taylor series approximation. Evaluation results demonstrate that optimizing radar and beam steering enhances secure communications while maintaining quality of service. Furthermore, the utilization of IRS effectively prevents passive eavesdropping in both line-of-sight (LOS) and non-LOS scenarios. © 1997-2012 IEEE.

关键词： Convex optimization

来源：评论

学校读者我要写书评

暂无评论

MetaIClone: A Robust Iris Clone Detection Mechanism for the Metaverse

MetaIClone: A Robust Iris Clone Detection Mechanism for the ...

引用

2024 IEEE International Conference on Modeling, Simulation and Intelligent Computing, MoSICom 2024

作者： Verma, Palak Kaur, Harkeerat Indian Institute of Technology Jammu Department of Computer Science and Engineering Jammu and Kashmir181221 India

ISBN: (纸本)9798331533311

Recent advancements in artificial intelligence have paved the way for the metaverse to become an imminent reality. Within this virtual world, people can buy and sell virtual assets, meet up with friends, purchase goods, and attend event, all while using biometric traits for secure and seamless authentication. Iris recognition has become a highly prevalent method for human authentication due to its robustness and effectiveness. Despite its widespread use, these systems are vulnerable to various security threats, particularly when attackers use biometric replicas to impersonate legitimate users-commonly known as biometric cloning. To counteract such cloning attempts, an anti-cloning subsystem is incorporated with the sensor, enabling intelligent differentiation between genuine and clone iris traits. In this research, we present MetaIClone, a hybrid iris clone detection (ICD) system designed for secure identification within the metaverse. MetaIClone integrates the powerful features of two handcrafted feature extractors namely histogram of oriented gradient (HoG) and binary statistical image feature (BSIF), with the extracted feature vectors subsequently classified using two distinct support vector machines (SVMs). The final decision is achieved through score-level fusion with a fine-tuned MobileNet V2 model. We evaluated MetaIClone using benchmark datasets, including iris liveness detection Notre Dame 2017 (LivDet ND17) and Notre Dame cosmetic lenses dataset 2015 (NDCLD-15). Additionally, we introduced a digital clone generation technique using a generative adversarial network (GAN) model and tested the system on this novel dataset. MetaIClone excels in known attack scenarios, achieving an average classification error rate (ACER) of 0. 0 2%, and performs well in unseen attack scenarios with an ACER of 0. 0 5%, delivering competitive results in crossdatabase evaluations. Overall, MetaIClone outperforms other ICD models, offering superior accuracy with high computational

关键词： Support vector machines

来源：评论

学校读者我要写书评

暂无评论

Hyperspectral image restoration using noise gradient and dual priors under mixed noise conditions

引用

CAAI Transactions on Intelligence technology 2025年第1期10卷 72-93页

作者： Hazique Aetesam Suman Kumar Maji V.B.Surya Prasath Computer Science and Engineering Birla Institute of Technology MesraBiharIndia Computer Science and Engineering Indian Institute of Technology PatnaBiharIndia Department of Computer Science University of CincinnatiCincinnatiOhioUSA

Images obtained from hyperspectral sensors provide information about the target area that extends beyond the visible portions of the electromagnetic ***,due to sensor limitations and imperfections during the image acquisition and transmission phases,noise is introduced into the acquired image,which can have a negative impact on downstream analyses such as classification,target tracking,and spectral *** in hyperspectral images(HSI)is modelled as a combination from several sources,including Gaussian/impulse noise,stripes,and *** HSI restoration method for such a mixed noise model is ***,a joint optimisation framework is proposed for recovering hyperspectral data corrupted by mixed Gaussian-impulse noise by estimating both the clean data as well as the sparse/impulse noise ***,a hyper-Laplacian prior is used along both the spatial and spectral dimensions to express sparsity in clean image ***,to model the sparse nature of impulse noise,anℓ_(1)−norm over the impulse noise gradient is *** the proposed methodology employs two distinct priors,the authors refer to it as the hyperspectral dual prior(HySpDualP)*** the best of authors'knowledge,this joint optimisation framework is the first attempt in this *** handle the non-smooth and nonconvex nature of the generalℓ_(p)−norm-based regularisation term,a generalised shrinkage/thresholding(GST)solver is ***,an efficient split-Bregman approach is used to solve the resulting optimisation *** results on synthetic data and real HSI datacube obtained from hyperspectral sensors demonstrate that the authors’proposed model outperforms state-of-the-art methods,both visually and in terms of various image quality assessment metrics.

关键词： hyper-laplacian prior hyperspectral images image restoration mixed noise variational approach

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：