检索结果-内蒙古大学图书馆

A Facial Expression Recognition Method Integrating Uncertainty Estimation and Active Learning

Computers, Materials & Continua 2024年第10期81卷 533-548页

作者： Yujian Wang Jianxun Zhang Renhao Sun School of Computer Science and Engineering Chongqing University of TechnologyChongqing400054China Data Space Research Institute of Hefei Comprehensive National Science Center Hefei230088China

The effectiveness of facial expression recognition(FER)algorithms hinges on the model’s quality and the availability of a substantial amount of labeled expression ***,labeling large datasets demands significant human,time,and financial *** active learning methods have mitigated the dependency on extensive labeled data,a cold-start problem persists in small to medium-sized expression recognition *** issue arises because the initial labeled data often fails to represent the full spectrum of facial expression *** paper introduces an active learning approach that integrates uncertainty estimation,aiming to improve the precision of facial expression recognition regardless of dataset scale *** method is divided into two primary ***,the model undergoes self-supervised pre-training using contrastive learning and uncertainty estimation to bolster its feature extraction ***,the model is fine-tuned using the prior knowledge obtained from the pre-training phase to significantly improve recognition *** the pretraining phase,the model employs contrastive learning to extract fundamental feature representations from the complete unlabeled *** features are then weighted through a self-attention mechanism with rank ***,data from the low-weighted set is relabeled to further refine the model’s feature extraction *** pre-trained model is then utilized in active learning to select and label information-rich samples more *** results demonstrate that the proposed method significantly outperforms existing approaches,achieving an improvement in recognition accuracy of 5.09%and 3.82%over the best existing active learning methods,Margin,and Least Confidence methods,respectively,and a 1.61%improvement compared to the conventional segmented active learning method.

关键词： Expression recognition active learning self-supervised learning uncertainty estimation

来源：评论

学校读者我要写书评

暂无评论

User acceptability of blockchain technology for enabling electronic health record exchange

引用

Journal of Systems and Information Technology 2023年第3期25卷 268-295页

作者： Baltruschat, Lilly Marie Jaiman, Vikas Urovi, Visara Faculty of Science and Engineering University College Maastricht Maastricht University Maastricht Netherlands Institute of Data Science Faculty of Science and Engineering Maastricht University Maastricht Netherlands

Purpose: Blockchain systems have been proposed as a solution for exchanging electronic health records (EHR) because they enable data sharing in decentralised networks. This paper aims to analyse the user acceptability of blockchain technology in enabling EHR exchange and to formulate practical implications for increasing user acceptability. Design/methodology/approach: A technology acceptance model [extended Unified Theory of Acceptance and Use of Technology (UTAUT) model] was used as a framework to measure the effects of 13 factors. The authors conducted a survey and analysed data from 214 participants using partial least square path modelling. Findings: The acceptance of blockchain for EHR sharing is positively influenced by performance expectancy, social influence and perceived trust. Effort expectancy and facilitating conditions do not influence acceptance. The UTAUT model explains the variance in acceptance at 58.4%. Self-efficacy influences effort expectancy, incentives influence facilitating conditions and security predicts perceived trust. Practical implications: Three implications are drawn: (1) Users need to clearly understand system’s purpose, functions, security mechanism and environmental impacts. (2) Users are incentivised to share health data via a blockchain solution if the technology offers personalising options and health information. (3) Health personnel can socially impact patients to use blockchain-based solutions. Originality/value: Studies have shown that blockchain technology is a valuable solution for exchanging EHR. The novelty of this work is to identify how and why patients may accept this emerging technology for EHR exchange. © 2023, Emerald Publishing Limited.

关键词： Electronic data interchange

来源：评论

学校读者我要写书评

暂无评论

SMART AGRICULTURE: INTEGRATING AIR QUALITY MONITORING WITH DEEP LEARNING FOR PROCESS OPTIMIZATION*

引用

Scalable Computing 2025年第3期26卷 1005-1016页

作者： Shobana, J. Subramanian, A Venkata Balamurugan, P. Perumal, Sivakumar Sankari, V. Eldho, K.J. Nareshkumar, R. Department of Data Science and Business Systems School of Computing College of Engineering and Technology SRM Institute of Science and Technology Tamilnadu Kattankulathur India Department of Computer Science and Engineering GITAM University Bengaluru India Department of Computer Science and Engineering J N N Institute of Enginering India Saveetha university Chennai India Department of Computer Science and Engineering Sasi Institute of Technology and Engineering Andhra Pradesh Tadepalligudem India Department of Artificial Intelligence and Data Science K.Ramakrishnan College of Engineering Kariyamanickam Rd Tamil Nadu Samayapuram India PG and Research Department of Computer Science Mary Matha Arts and Science College Kerala Mananthavady India Department of Networking and Communications School of Computing College of Engineering and Technology SRM Institute of Science and Technology Tamilnadu Kattankulathur India

Modernization and intense industrialization have led to a substantial improvement in people’s quality of life. However, the aspiration for achieving an improved quality of life results in environmental contamination. A primary consequence of environmental degradation is air pollution, resulting from rising levels of poisonous chemicals in the atmosphere, which may induce detrimental health conditions in humans. It is harmful to both humans and agriculture. Given that the effects of air pollution on plants may not be readily apparent, it is important to analyse the necessary data and compute the outcomes. Farmers prioritise on pests and plant diseases, frequently neglecting the detrimental impacts of air pollution. Some plant species can withstand high amounts of pollution from suspended particulate matter and accumulated gases, while others are more susceptible to harm. Therefore, plants’ reaction to air pollution is influenced by the kind of harmful compounds, their levels, and the plant’s susceptibility to them. The LSTM +CNN Proposed Ensemble method may be used to analyse the impact of air pollution on agriculture by examining trends in crop production over time and predicting which crop is more resistant based on the pollution data. The initiative created for this aim may assist farmers in determining the most suitable crop to cultivate in their fields to minimize the impact of air pollution on agricultural yield. The findings show deep learning algorithms correctly predict hourly pollutant concentrations such as carbon monoxide, sulphur dioxide, nitrogen dioxide, ground-level ozone, and particulate matter 2.5, along with the hourly Air Quality Index (AQI) for California. A proposed model used test RMSE values as a measure to evaluate prediction performance, achieving the best possible results. © 2025 SCPE.

关键词： Nitrogen oxides

来源：评论

学校读者我要写书评

暂无评论

Face Anti-Spoofing with Unknown Attacks:A Comprehensive Feature Extraction and Representation Perspective

引用

Journal of Computer science & Technology 2024年第4期39卷 827-840页

作者： Li-Min Li Bin-Wu Wang Xu Wang Peng-Kun Wang Yu-Dong Zhang Yang Wang School of Software Engineering University of Science and Technology of ChinaHefei 230000China School of Data Science University of Science and Technology of ChinaHefei 23000China Suzhou Institute for Advanced Research University of Science and Technology of ChinaSuzhou 215000China

Face anti-spoofing aims at detecting whether the input is a real photo of a user(living)or a fake(spoofing)*** new types of attacks keep emerging,the detection of unknown attacks,known as Zero-Shot Face Anti-Spoofing(ZSFA),has become increasingly important in both academia and *** ZSFA methods mainly focus on extracting discriminative features between spoofing and living ***,the nature of the spoofing faces is to trick anti-spoofing systems by mimicking the livings,therefore the deceptive features between the known attacks and the livings,which have been ignored by existing ZSFA methods,are essential to comprehensively represent the ***,existing ZSFA models are incapable of learning the complete representations of living faces and thus fall short of effectively detecting newly emerged *** tackle this problem,we propose an innovative method that effectively captures both the deceptive and discriminative features distinguishing between genuine and spoofing *** method consists of two main components:a two-against-all training strategy and a semantic *** two-against-all training strategy is employed to separate deceptive and discriminative *** address the subsequent invalidation issue of categorical functions and the dominance disequilibrium issue among different dimensions of features after importing deceptive features,we introduce a modified semantic *** autoencoder is designed to map all extracted features to a semantic space,thereby achieving a balance in the dominance of each feature *** combine our method with the feature extraction model ResNet50,and experimental results show that the trained ResNet50 model simultaneously achieves a feasible detection of unknown attacks and comparably accurate detection of known *** results confirm the superiority and effectiveness of our proposed method in identifying the living with the interference of both known

关键词： face anti-spoofing spoof detection zero-shot learning convolutional neural network deep learning

来源：评论

学校读者我要写书评

暂无评论

Classification and structural characteristics of amorphous materials based on interpretable deep learning

引用

Chinese Physics B 2023年第9期32卷 356-363页

作者：崔佳梅李韵洁赵偲郑文 Institute of Public Safety and Big Data College of Data ScienceTaiyuan University of TechnologyJinzhong 030600China Shanxi Engineering Research Center for Intelligent Data Assisted Treatment Changzhi Medical CollegeChangzhi 046000China

Defining the structure characteristics of amorphous materials is one of the fundamental problems that need to be solved urgently in complex materials because of their complex structure and long-range *** this study,we develop an interpretable deep learning model capable of accurately classifying amorphous configurations and characterizing their structural *** results demonstrate that the multi-dimensional hybrid convolutional neural network can classify the two-dimensional(2D)liquids and amorphous solids of molecular dynamics *** classification process does not make a priori assumptions on the amorphous particle environment,and the accuracy is 92.75%,which is better than other convolutional neural ***,our model utilizes the gradient-weighted activation-like mapping method,which generates activation-like heat maps that can precisely identify important structures in the amorphous configuration *** obtain an order parameter from the heatmap and conduct finite scale analysis of this *** findings demonstrate that the order parameter effectively captures the amorphous phase transition process across various *** results hold significant scientific implications for the study of amorphous structural characteristics via deep learning.

关键词： amorphous interpretable deep learning image classification finite scale analysis

来源：评论

学校读者我要写书评

暂无评论

Autonomous Recommendation of Fault Detection Algorithms for Spacecraft

引用

IEEE/CAA Journal of Automatica Sinica 2024年第1期11卷 273-275页

作者： Wenbo Li Baoling Ning the National Key Laboratory of Science and Technology on Space Intelligent Control Beijing Institute of Control EngineeringBeijing 100094China the School of Data Science and Technology Heilongjiang UniversityHarbin 150001China

Dear Editor, This letter deals with the problem of algorithm recommendation for online fault detection of spacecraft. By transforming the time series data into distributions and introducing a distribution-aware measure, a principal method is designed for quantifying the detectabilities of fault detection algorithms over special datasets.

关键词： fault introducing letter

来源：评论

学校读者我要写书评

暂无评论

Enhancing wireless sensing via a target-mounted intelligent reflecting surface

引用

National science Review 2023年第8期10卷 13-14页

作者： Xiaodan Shao Rui Zhang School of Science and Engineering Shenzhen Research Institute of Big Data The Chinese University of Hong Kong Department of Electrical and Computer Engineering National University of Singapore

An intelligent reflecting surface(IRS),or its various equivalents such as an reconfigurable intelligent surface(RIS), is an emerging technology to control radio signal propagation in wireless systems. An IRS is a digitally controlled metasurface consisting of a large number of passive reflecting elements, which are connected to a smart controller to enable dynamic adjustments of the amplitude and/or phase of the incident signal on each element independently [1].

关键词：

来源：评论

学校读者我要写书评

暂无评论

Are Object Recognition Models Effective and Unbiased for Biometric Recognition? 18

Are Object Recognition Models Effective and Unbiased for Bio...

引用

18th IEEE International Joint Conference on Biometrics, IJCB 2024

作者： Kumar, Vishesh Agarwal, Akshay Indian Institute of Science Education and Research Department of Data Science and Engineering Bhopal India

ISBN: (纸本)9798350364132

Can the general-purpose pre-trained image classifier be effective for biometric recognition? Due to the prevalence of generic features in these models, we assert that they can be used to extract the discriminating feature of the biometrics images which help encode the object images. The utilization of features from the pre-trained model can significantly limit the training requirements and can be deployed on computationally limited devices. In this research, we have conducted an extensive study to perform biometric recognition using existing state-of-the-art pre-trained deep neural networks. The experiments are not limited to any modality or imaging spectrum to ensure the findings are trustworthy and thorough. The experiments conducted on various modalities including multi-modal biometric recognition as well showcase that the pre-trained deep networks can be a suitable option for cost-effective biometric recognition. Further, we have also evaluated the fusion of deep neural networks to see if the performance can further be improved without significantly increasing the computational cost. © 2024 IEEE.

关键词： Biometrics

来源：评论

学校读者我要写书评

暂无评论

Benchmarking In-the-wild Soft Biometric Attribute Identification 18

Benchmarking In-the-wild Soft Biometric Attribute Identifica...

引用

18th IEEE International Joint Conference on Biometrics, IJCB 2024

作者： Manju, R.A. Agarwal, Akshay Indian Institute of Science Education and Research Department of Data Science and Engineering Bhopal India

ISBN: (数字)9798350364132

ISBN: (纸本)9798350364132

The active involvement of different demographic entities, whether positive or negative, demands effective identification of diverse populations. One way to quickly perform this identification is by segregating individuals based on their soft biometric attributes such as race and gender. Further, the soft biometric recognition technology has the potential to enhance individual privacy by providing a less invasive identification mode than conventional biometric methods. For example, a specific disbursement of government benefits might not require the exact identity of a person but only soft biometric attributes for its actual distribution. Secondly, the correct identification of race and gender can restrict the identity search space in which we need to look for an imposter. However, despite tremendous literature on identity recognition, limited work has been done to identify soft biometric attributes in challenging evaluation settings accurately. Therefore, in this research, for the first time, we have performed extensive experiments to benchmark the effectiveness of pure convolutional networks and attention networks to identify gender and ethnicity in the in-distribution and out-of-distribution (OOD) settings. We employed diverse face datasets to benchmark our methods, including UTKFace and FairFace. In contrast to the general understanding that OOD images lead to poor performance, we observe significant performance in identifying soft biometric attributes, including race and gender classification. © 2024 IEEE.

关键词： Differential privacy

来源：评论

学校读者我要写书评

暂无评论

A Natural Language Processing Based Framework for Early Detection of Anorexia via Sequential Text Processing 25

A Natural Language Processing Based Framework for Early Dete...

引用

25th Working Notes of the Conference and Labs of the Evaluation Forum, CLEF 2024

作者： Sarangi, Prateek Kumar, Sumit Agarwal, Shraddha Basu, Tanmay Department of Data Science and Engineering Indian Institute of Science Education and Research Bhopal India

Task 2 of eRisk shared tasks in CLEF 2024 aims to develop text mining solutions for early prediction of anorexia using sequentially posted texts over social media. Anorexia is an eating disorder, a kind of mental illness, where people have distorted perceptions of their body weights and accordingly manipulate their food habits, which often results in deficient body weight. The aim here is to identify anorexia by processing the interactions over social media of an individual through text mining models. The organisers provided training data with ground truths for training the models and test data for evaluating their performance. The BioNLP research group at the Indian institute of science Education and Research Bhopal (IISERB) participated in Task 2 and submitted five runs for five text mining frameworks. Five different classifiers and individual feature engineering techniques were used to develop frameworks. The bag-of-words model and transformer-based embedding were used as features for individual classifiers. The performance of multiple classifiers was evaluated using the training corpus. Then Random Forest, Adaptive Boosting, Logistic Regression, Support Vector Machine (SVM), and Longformer classifiers were chosen to run on the test set. Experimental results show that SVM and AdaBoost classifiers using the TF-IDF-based weighting schemes achieved the highest precision score among all submitted runs in Task 2 of eRisk 2024. However, the performance of our models in terms of the other metrics, like recall, f-score, ERDE, etc., is not reasonably good compared to the other runs. Hence, we plan to develop transformer-based embeddings from scratch using data collected from multiple social media platforms. © 2024 Copyright for this paper by its authors.

关键词： Adaptive boosting

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：