Pub Date : 2024-09-11DOI: 10.3390/electronics13183612
Paulo H. N. Gonçalves, Hendrio Bragança, Eduardo Souto
Mobile and wearable devices have revolutionized the field of continuous user activity monitoring. However, analyzing the vast and intricate data captured by the sensors of these devices poses significant challenges. Deep neural networks have shown remarkable accuracy in Human Activity Recognition (HAR), but their application on mobile and wearable devices is constrained by limited computational resources. To address this limitation, we propose a novel method called Knowledge Distillation for Human Activity Recognition (KD-HAR) that leverages the knowledge distillation technique to compress deep neural network models for HAR using inertial sensor data. Our approach transfers the acquired knowledge from high-complexity teacher models (state-of-the-art models) to student models with reduced complexity. This compression strategy allows us to maintain performance while keeping computational costs low. To assess the compression capabilities of our approach, we evaluate it using two popular databases (UCI-HAR and WISDM) comprising inertial sensor data from smartphones. Our results demonstrate that our method achieves competitive accuracy, even at compression rates ranging from 18 to 42 times the number of parameters compared to the original teacher model.
{"title":"Efficient Human Activity Recognition on Wearable Devices Using Knowledge Distillation Techniques","authors":"Paulo H. N. Gonçalves, Hendrio Bragança, Eduardo Souto","doi":"10.3390/electronics13183612","DOIUrl":"https://doi.org/10.3390/electronics13183612","url":null,"abstract":"Mobile and wearable devices have revolutionized the field of continuous user activity monitoring. However, analyzing the vast and intricate data captured by the sensors of these devices poses significant challenges. Deep neural networks have shown remarkable accuracy in Human Activity Recognition (HAR), but their application on mobile and wearable devices is constrained by limited computational resources. To address this limitation, we propose a novel method called Knowledge Distillation for Human Activity Recognition (KD-HAR) that leverages the knowledge distillation technique to compress deep neural network models for HAR using inertial sensor data. Our approach transfers the acquired knowledge from high-complexity teacher models (state-of-the-art models) to student models with reduced complexity. This compression strategy allows us to maintain performance while keeping computational costs low. To assess the compression capabilities of our approach, we evaluate it using two popular databases (UCI-HAR and WISDM) comprising inertial sensor data from smartphones. Our results demonstrate that our method achieves competitive accuracy, even at compression rates ranging from 18 to 42 times the number of parameters compared to the original teacher model.","PeriodicalId":11646,"journal":{"name":"Electronics","volume":null,"pages":null},"PeriodicalIF":2.9,"publicationDate":"2024-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142209067","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
With the rapid development of modern power systems, the structure and operation of distribution networks are becoming increasingly complex, demanding higher levels of intelligence and digitization. Digital twin, as a virtual cutting-edge technique, can effectively reflect the operational status of distribution networks, offering new possibilities for real-time monitoring, optimization and other functions for distribution networks. Building efficient and accurate models is the foundation of enabling a digital twin of distribution networks. This paper proposes a digital twin operating system for distribution networks with renewable energy based on robust state estimation and deep learning-based renewable energy prediction. Furthermore, the identification and correction of possible bad or missing data based on deep learning are also included to purify the input data for the digital twin system. A digital twin test platform is also proposed in the paper. A case study and evaluations based on a real-time digital simulator are carried out to verify the accuracy and real-time performance of the established digital twin system. In general, the proposed method can provide the basis and foundation for distribution network management and operation, as well as intelligent power system operation.
{"title":"Digital Twin for Modern Distribution Networks by Improved State Estimation with Consideration of Bad Date Identification","authors":"Huiqiang Zhi, Rui Mao, Longfei Hao, Xiao Chang, Xiangyu Guo, Liang Ji","doi":"10.3390/electronics13183613","DOIUrl":"https://doi.org/10.3390/electronics13183613","url":null,"abstract":"With the rapid development of modern power systems, the structure and operation of distribution networks are becoming increasingly complex, demanding higher levels of intelligence and digitization. Digital twin, as a virtual cutting-edge technique, can effectively reflect the operational status of distribution networks, offering new possibilities for real-time monitoring, optimization and other functions for distribution networks. Building efficient and accurate models is the foundation of enabling a digital twin of distribution networks. This paper proposes a digital twin operating system for distribution networks with renewable energy based on robust state estimation and deep learning-based renewable energy prediction. Furthermore, the identification and correction of possible bad or missing data based on deep learning are also included to purify the input data for the digital twin system. A digital twin test platform is also proposed in the paper. A case study and evaluations based on a real-time digital simulator are carried out to verify the accuracy and real-time performance of the established digital twin system. In general, the proposed method can provide the basis and foundation for distribution network management and operation, as well as intelligent power system operation.","PeriodicalId":11646,"journal":{"name":"Electronics","volume":null,"pages":null},"PeriodicalIF":2.9,"publicationDate":"2024-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142209065","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-11DOI: 10.3390/electronics13183605
An Zhao, Wenzhong Yang, Danny Chen, Fuyuan Wei
Remote-sensing image captioning (RSIC) aims to generate descriptive sentences for ages by capturing both local and global semantic information. This task is challenging due to the diverse object types and varying scenes in ages. To address these challenges, we propose a positional-channel semantic fusion transformer (PCSFTr). The PCSFTr model employs scene classification to initially extract visual features and learn semantic information. A novel positional-channel multi-headed self-attention (PCMSA) block captures spatial and channel dependencies simultaneously, enriching the semantic information. The feature fusion (FF) module further enhances the understanding of semantic relationships. Experimental results show that PCSFTr significantly outperforms existing methods. Specifically, the BLEU-4 index reached 78.42% in UCM-caption, 54.42% in RSICD, and 69.01% in NWPU-captions. This research provides new insights into RSIC by offering a more comprehensive understanding of semantic information and relationships within images and improving the performance of image captioning models.
{"title":"Enhanced Transformer for Remote-Sensing Image Captioning with Positional-Channel Semantic Fusion","authors":"An Zhao, Wenzhong Yang, Danny Chen, Fuyuan Wei","doi":"10.3390/electronics13183605","DOIUrl":"https://doi.org/10.3390/electronics13183605","url":null,"abstract":"Remote-sensing image captioning (RSIC) aims to generate descriptive sentences for ages by capturing both local and global semantic information. This task is challenging due to the diverse object types and varying scenes in ages. To address these challenges, we propose a positional-channel semantic fusion transformer (PCSFTr). The PCSFTr model employs scene classification to initially extract visual features and learn semantic information. A novel positional-channel multi-headed self-attention (PCMSA) block captures spatial and channel dependencies simultaneously, enriching the semantic information. The feature fusion (FF) module further enhances the understanding of semantic relationships. Experimental results show that PCSFTr significantly outperforms existing methods. Specifically, the BLEU-4 index reached 78.42% in UCM-caption, 54.42% in RSICD, and 69.01% in NWPU-captions. This research provides new insights into RSIC by offering a more comprehensive understanding of semantic information and relationships within images and improving the performance of image captioning models.","PeriodicalId":11646,"journal":{"name":"Electronics","volume":null,"pages":null},"PeriodicalIF":2.9,"publicationDate":"2024-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142209057","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-11DOI: 10.3390/electronics13183616
Viorel-Costin Banța, Ștefan Bunea, Daniela Țuțui, Raluca Florentina Crețu
Higher education institutions are increasingly concerned with providing students with sustainable education by developing the necessary competencies for various roles in the business environment. To be more effective, courses must develop technological, organizational and environmental (TOE) competencies in an integrated manner. SAP is a tool that yields this possibility through the diversity of IT solutions by ensuring a significant increase in employability rates. Learning SAP is a competitive advantage because it helps with all aspects of digital transformation within the concept of Industry 4.0. Our research aims to investigate to what extent students perceive that they have acquired the knowledge and competencies specific to the three dimensions of the TOE framework within the SAP course. We have added a fourth dimension to the TOE framework: the learning context (L) considering the impact of the educational environment on perceived learning outcomes. Data collection was based on a questionnaire distributed to students enrolled in the SAP course in the academic year 2023–2024 at Bucharest University of Economic Studies (BUES). The data were processed using correlation and regression analysis. Reconfiguring the content elements of SAP courses based on the TOE framework would ensure greater effectiveness in the learning process.
高等教育机构越来越关注通过培养学生在商业环境中扮演各种角色所需的能力，为学生提供可持续教育。为了提高效率，课程必须以综合方式培养技术、组织和环境（TOE）能力。SAP 是一种工具，可通过信息技术解决方案的多样性实现这种可能性，确保显著提高就业率。学习 SAP 是一种竞争优势，因为它有助于实现工业 4.0 概念中数字化转型的各个方面。我们的研究旨在调查学生在多大程度上认为他们已经掌握了 SAP 课程中 TOE 框架三个维度所特有的知识和能力。考虑到教育环境对感知学习成果的影响，我们在 TOE 框架中增加了第四个维度：学习环境（L）。数据收集基于向布加勒斯特经济研究大学（BUES）2023-2024 学年 SAP 课程学生发放的调查问卷。数据处理采用了相关分析和回归分析。根据 TOE 框架重新配置 SAP 课程的内容要素将确保学习过程更加有效。
{"title":"Challenges in Information Systems Curricula: Effectiveness of Systems Application Products in Data Processing Learning in Higher Education through a Technological, Organizational and Environmental Framework","authors":"Viorel-Costin Banța, Ștefan Bunea, Daniela Țuțui, Raluca Florentina Crețu","doi":"10.3390/electronics13183616","DOIUrl":"https://doi.org/10.3390/electronics13183616","url":null,"abstract":"Higher education institutions are increasingly concerned with providing students with sustainable education by developing the necessary competencies for various roles in the business environment. To be more effective, courses must develop technological, organizational and environmental (TOE) competencies in an integrated manner. SAP is a tool that yields this possibility through the diversity of IT solutions by ensuring a significant increase in employability rates. Learning SAP is a competitive advantage because it helps with all aspects of digital transformation within the concept of Industry 4.0. Our research aims to investigate to what extent students perceive that they have acquired the knowledge and competencies specific to the three dimensions of the TOE framework within the SAP course. We have added a fourth dimension to the TOE framework: the learning context (L) considering the impact of the educational environment on perceived learning outcomes. Data collection was based on a questionnaire distributed to students enrolled in the SAP course in the academic year 2023–2024 at Bucharest University of Economic Studies (BUES). The data were processed using correlation and regression analysis. Reconfiguring the content elements of SAP courses based on the TOE framework would ensure greater effectiveness in the learning process.","PeriodicalId":11646,"journal":{"name":"Electronics","volume":null,"pages":null},"PeriodicalIF":2.9,"publicationDate":"2024-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142209071","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-11DOI: 10.3390/electronics13183601
Brunel Rolack Kikissagbe, Meddi Adda
The rise of the Internet of Things (IoT) has transformed our daily lives by connecting objects to the Internet, thereby creating interactive, automated environments. However, this rapid expansion raises major security concerns, particularly regarding intrusion detection. Traditional intrusion detection systems (IDSs) are often ill-suited to the dynamic and varied networks characteristic of the IoT. Machine learning is emerging as a promising solution to these challenges, offering the intelligence and flexibility needed to counter complex and evolving threats. This comprehensive review explores different machine learning approaches for intrusion detection in IoT systems, covering supervised, unsupervised, and deep learning methods, as well as hybrid models. It assesses their effectiveness, limitations, and practical applications, highlighting the potential of machine learning to enhance the security of IoT systems. In addition, the study examines current industry issues and trends, highlighting the importance of ongoing research to keep pace with the rapidly evolving IoT security ecosystem.
{"title":"Machine Learning-Based Intrusion Detection Methods in IoT Systems: A Comprehensive Review","authors":"Brunel Rolack Kikissagbe, Meddi Adda","doi":"10.3390/electronics13183601","DOIUrl":"https://doi.org/10.3390/electronics13183601","url":null,"abstract":"The rise of the Internet of Things (IoT) has transformed our daily lives by connecting objects to the Internet, thereby creating interactive, automated environments. However, this rapid expansion raises major security concerns, particularly regarding intrusion detection. Traditional intrusion detection systems (IDSs) are often ill-suited to the dynamic and varied networks characteristic of the IoT. Machine learning is emerging as a promising solution to these challenges, offering the intelligence and flexibility needed to counter complex and evolving threats. This comprehensive review explores different machine learning approaches for intrusion detection in IoT systems, covering supervised, unsupervised, and deep learning methods, as well as hybrid models. It assesses their effectiveness, limitations, and practical applications, highlighting the potential of machine learning to enhance the security of IoT systems. In addition, the study examines current industry issues and trends, highlighting the importance of ongoing research to keep pace with the rapidly evolving IoT security ecosystem.","PeriodicalId":11646,"journal":{"name":"Electronics","volume":null,"pages":null},"PeriodicalIF":2.9,"publicationDate":"2024-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142209059","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-11DOI: 10.3390/electronics13183609
Hong Zhang, Kunzhong Miao, Huangzhi Yu, Yifeng Niu
The existing task assignment algorithms usually solve only a point-based model. This paper proposes a novel algorithm for task assignment in detection search tasks. Firstly, the optimal reconnaissance path is generated by considering the drone’s position and attitude information, as well as the type of heterogeneous targets present in the actual scene. Subsequently, an adaptive crowding distance calculation (ACD-NSGA-II) is proposed based on the relative position of solutions in space, taking into account the spatial distribution of parent solutions and constraints imposed by uncertain targets and terrain. Finally, comparative experiments using digital simulation are conducted under two different target probability scenarios. Moreover, the improved algorithm is further evaluated across 100 cases, and a comparison of the Pareto solution set with other algorithms is conducted to demonstrate the algorithm’s overall adaptability.
{"title":"Multi-UAV Reconnaissance Task Assignment for Heterogeneous Targets with ACD-NSGA-II Algorithm","authors":"Hong Zhang, Kunzhong Miao, Huangzhi Yu, Yifeng Niu","doi":"10.3390/electronics13183609","DOIUrl":"https://doi.org/10.3390/electronics13183609","url":null,"abstract":"The existing task assignment algorithms usually solve only a point-based model. This paper proposes a novel algorithm for task assignment in detection search tasks. Firstly, the optimal reconnaissance path is generated by considering the drone’s position and attitude information, as well as the type of heterogeneous targets present in the actual scene. Subsequently, an adaptive crowding distance calculation (ACD-NSGA-II) is proposed based on the relative position of solutions in space, taking into account the spatial distribution of parent solutions and constraints imposed by uncertain targets and terrain. Finally, comparative experiments using digital simulation are conducted under two different target probability scenarios. Moreover, the improved algorithm is further evaluated across 100 cases, and a comparison of the Pareto solution set with other algorithms is conducted to demonstrate the algorithm’s overall adaptability.","PeriodicalId":11646,"journal":{"name":"Electronics","volume":null,"pages":null},"PeriodicalIF":2.9,"publicationDate":"2024-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142209061","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Knowledge graphs equipped with graph network networks (GNNs) have led to a successful step forward in alleviating cold start problems in recommender systems. However, the performance highly depends on precious high-quality knowledge graphs and supervised labels. This paper argues that existing knowledge-graph-based recommendation methods still suffer from insufficiently exploiting sparse information and the mismatch between personalized interests and general knowledge. This paper proposes a model named Adaptive Knowledge Contrastive Learning with Dynamic Attention (AKCL-DA) to address the above challenges. Specifically, instead of building contrastive views by randomly discarding information, in this study, an adaptive data augmentation method was designed to leverage sparse information effectively. Furthermore, a personalized dynamic attention network was proposed to capture knowledge-aware personalized behaviors by dynamically adjusting user attention, therefore alleviating the mismatch between personalized behavior and general knowledge. Extensive experiments on Yelp2018, LastFM, and MovieLens datasets show that AKCL-DA achieves a strong performance, improving the NDCG by 4.82%, 13.66%, and 4.41% compared to state-of-the-art models, respectively.
{"title":"Adaptive Knowledge Contrastive Learning with Dynamic Attention for Recommender Systems","authors":"Hongchan Li, Jinming Zheng, Baohua Jin, Haodong Zhu","doi":"10.3390/electronics13183594","DOIUrl":"https://doi.org/10.3390/electronics13183594","url":null,"abstract":"Knowledge graphs equipped with graph network networks (GNNs) have led to a successful step forward in alleviating cold start problems in recommender systems. However, the performance highly depends on precious high-quality knowledge graphs and supervised labels. This paper argues that existing knowledge-graph-based recommendation methods still suffer from insufficiently exploiting sparse information and the mismatch between personalized interests and general knowledge. This paper proposes a model named Adaptive Knowledge Contrastive Learning with Dynamic Attention (AKCL-DA) to address the above challenges. Specifically, instead of building contrastive views by randomly discarding information, in this study, an adaptive data augmentation method was designed to leverage sparse information effectively. Furthermore, a personalized dynamic attention network was proposed to capture knowledge-aware personalized behaviors by dynamically adjusting user attention, therefore alleviating the mismatch between personalized behavior and general knowledge. Extensive experiments on Yelp2018, LastFM, and MovieLens datasets show that AKCL-DA achieves a strong performance, improving the NDCG by 4.82%, 13.66%, and 4.41% compared to state-of-the-art models, respectively.","PeriodicalId":11646,"journal":{"name":"Electronics","volume":null,"pages":null},"PeriodicalIF":2.9,"publicationDate":"2024-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142209054","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-10DOI: 10.3390/electronics13183599
Suyao Liu, Chunmei Xu, Yifei Zhang, Haoying Pei, Kan Dong, Ning Yang, Yingtao Ma
Conventional methods of parameterizing fuel cell hybrid power systems (FCHPS) often rely on engineering experience, which leads to problems such as increased economic costs and excessive weight of the system. These shortcomings limit the performance of FCHPS in real-world applications. To address these issues, this paper proposes a novel method for optimizing the parameter configuration of FCHPS. First, the power and energy requirements of the vehicle are determined through traction calculations, and a real-time energy management strategy is used to ensure efficient power distribution. On this basis, a multi-objective parameter configuration optimization model is developed, which comprehensively considers economic cost and system weight, and uses a particle swarm optimization (PSO) algorithm to determine the optimal configuration of each power source. The optimization results show that the system economic cost is reduced by 8.76% and 18.05% and the weight is reduced by 11.47% and 9.13%, respectively, compared with the initial configuration. These results verify the effectiveness of the proposed optimization strategy and demonstrate its potential to improve the overall performance of the FCHPS.
{"title":"Multi-Objective Parameter Configuration Optimization of Hydrogen Fuel Cell Hybrid Power System for Locomotives","authors":"Suyao Liu, Chunmei Xu, Yifei Zhang, Haoying Pei, Kan Dong, Ning Yang, Yingtao Ma","doi":"10.3390/electronics13183599","DOIUrl":"https://doi.org/10.3390/electronics13183599","url":null,"abstract":"Conventional methods of parameterizing fuel cell hybrid power systems (FCHPS) often rely on engineering experience, which leads to problems such as increased economic costs and excessive weight of the system. These shortcomings limit the performance of FCHPS in real-world applications. To address these issues, this paper proposes a novel method for optimizing the parameter configuration of FCHPS. First, the power and energy requirements of the vehicle are determined through traction calculations, and a real-time energy management strategy is used to ensure efficient power distribution. On this basis, a multi-objective parameter configuration optimization model is developed, which comprehensively considers economic cost and system weight, and uses a particle swarm optimization (PSO) algorithm to determine the optimal configuration of each power source. The optimization results show that the system economic cost is reduced by 8.76% and 18.05% and the weight is reduced by 11.47% and 9.13%, respectively, compared with the initial configuration. These results verify the effectiveness of the proposed optimization strategy and demonstrate its potential to improve the overall performance of the FCHPS.","PeriodicalId":11646,"journal":{"name":"Electronics","volume":null,"pages":null},"PeriodicalIF":2.9,"publicationDate":"2024-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142209055","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-09DOI: 10.3390/electronics13173580
Yizhou Mao, Shubin Zheng, Liming Li, Renjie Shi, Xiaoxue An
Rail surface defect detection is vital for railway safety. Traditional methods falter with varying defect sizes and complex backgrounds, while two-stage deep learning models, though accurate, lack real-time capabilities. To overcome these challenges, we propose an enhanced one-stage detection model based on CenterNet. We replace ResNet with ResNeXt and implement a multi-branch structure for better low-level feature extraction. Additionally, we integrate SKNet attention mechanism with the C2f structure from YOLOv8, improving the model’s focus on critical image regions and enhancing the detection of minor defects. We also introduce an elliptical Gaussian kernel for size regression loss, better representing the aspect ratio of rail defects. This approach enhances detection accuracy and speeds up training. Our model achieves a mean accuracy (mAP) of 0.952 on the rail defects dataset, outperforming other models with a 6.6% improvement over the original and a 35.5% increase in training speed. These results demonstrate the efficiency and reliability of our method for rail defect detection.
{"title":"Research on Rail Surface Defect Detection Based on Improved CenterNet","authors":"Yizhou Mao, Shubin Zheng, Liming Li, Renjie Shi, Xiaoxue An","doi":"10.3390/electronics13173580","DOIUrl":"https://doi.org/10.3390/electronics13173580","url":null,"abstract":"Rail surface defect detection is vital for railway safety. Traditional methods falter with varying defect sizes and complex backgrounds, while two-stage deep learning models, though accurate, lack real-time capabilities. To overcome these challenges, we propose an enhanced one-stage detection model based on CenterNet. We replace ResNet with ResNeXt and implement a multi-branch structure for better low-level feature extraction. Additionally, we integrate SKNet attention mechanism with the C2f structure from YOLOv8, improving the model’s focus on critical image regions and enhancing the detection of minor defects. We also introduce an elliptical Gaussian kernel for size regression loss, better representing the aspect ratio of rail defects. This approach enhances detection accuracy and speeds up training. Our model achieves a mean accuracy (mAP) of 0.952 on the rail defects dataset, outperforming other models with a 6.6% improvement over the original and a 35.5% increase in training speed. These results demonstrate the efficiency and reliability of our method for rail defect detection.","PeriodicalId":11646,"journal":{"name":"Electronics","volume":null,"pages":null},"PeriodicalIF":2.9,"publicationDate":"2024-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142209110","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-09-09DOI: 10.3390/electronics13173577
Ye Jin Kim, Jung Ho Song, Ki Hwan Cho, Jong Hyeon Shin, Jong Sik Kim, Jung Sik Yoon, Sang Jeen Hong
Existing etch endpoint detection (EPD) methods, primarily based on single wavelengths, have limitations, such as low signal-to-noise ratios and the inability to consider the long-term dependencies of time series data. To address these issues, this study proposes a context of time series data using long short-term memory (LSTM), a kind of recurrent neural network (RNN). The proposed method is based on the time series data collected through optical emission spectroscopy (OES) data during the SiO2 etching process. After training the LSTM model, the proposed method demonstrated the ability to detect the etch endpoint more accurately than existing methods by considering the entire time series. The LSTM model achieved an accuracy of 97.1% in a given condition, which shows that considering the flow and context of time series data can significantly reduce the false detection rate. To improve the performance of the proposed LSTM model, we created an attention-based LSTM model and confirmed that the model accuracy is 98.2%, and the performance is improved compared to that of the existing LSTM model.
{"title":"Improved Plasma Etch Endpoint Detection Using Attention-Based Long Short-Term Memory Machine Learning","authors":"Ye Jin Kim, Jung Ho Song, Ki Hwan Cho, Jong Hyeon Shin, Jong Sik Kim, Jung Sik Yoon, Sang Jeen Hong","doi":"10.3390/electronics13173577","DOIUrl":"https://doi.org/10.3390/electronics13173577","url":null,"abstract":"Existing etch endpoint detection (EPD) methods, primarily based on single wavelengths, have limitations, such as low signal-to-noise ratios and the inability to consider the long-term dependencies of time series data. To address these issues, this study proposes a context of time series data using long short-term memory (LSTM), a kind of recurrent neural network (RNN). The proposed method is based on the time series data collected through optical emission spectroscopy (OES) data during the SiO2 etching process. After training the LSTM model, the proposed method demonstrated the ability to detect the etch endpoint more accurately than existing methods by considering the entire time series. The LSTM model achieved an accuracy of 97.1% in a given condition, which shows that considering the flow and context of time series data can significantly reduce the false detection rate. To improve the performance of the proposed LSTM model, we created an attention-based LSTM model and confirmed that the model accuracy is 98.2%, and the performance is improved compared to that of the existing LSTM model.","PeriodicalId":11646,"journal":{"name":"Electronics","volume":null,"pages":null},"PeriodicalIF":2.9,"publicationDate":"2024-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142209105","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：