检索结果-内蒙古大学图书馆

Integrating cycleGAN and BERT for Chinese text style transfer

Multimedia Tools and Applications 2024年 1-20页

作者： Chou, Chien-Hsing Chou, Cheng-Hou Hsieh, Yi-Zeng Yang, Tzu-Shien Department of Mathematics and Information Education National Taipei University of Education Taipei Taiwan Department of Electrical Engineering National Taiwan University of Science and Technology Taipei Taiwan Department of Electrical and Computer Engineering Tamkang University New Taipei Taiwan

In this study, we integrate the Bidirectional Encoder Representations from Transformers (BERT) model with the Cycle Generative Adversarial Network (CycleGAN) to create a system for Chinese text style transfer. Natural language processing (NLP) involves converting human languages into data interpretable by computers, enabling applications like text classification, chatbots, and dialogue systems. Recent advancements, such as Google's transformer model and the BERT technique, have significantly improved NLP capabilities through self-attention mechanisms and unsupervised pretraining. Text style transfer modifies the style of texts without altering their semantics. Previous methods like StyIns and models based on disentangled representation learning highlight the challenges of retaining text meaning during style transfer. Our system leverages CycleGAN’s unsupervised learning to convert unpaired data between wuxia and fantasy styles while preserving semantics. Using the pretrained BERT model from the Chinese Knowledge and Information Processing (CKIP) Lab, our experimental results demonstrate successful style conversion, maintaining the original meanings of texts. This integration of BERT and CycleGAN shows promise for further advancements in NLP applications. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

A generative model for synchrophasor network traffic in smart grids 59

A generative model for synchrophasor network traffic in smar...

引用

59th Annual Conference on Information sciences and Systems, CISS 2025

作者： Kosta, Leo Leger, Aaron St. United States Military Academy Department of Electrical Engineering and Computer Science NY United States

ISBN: (纸本)9798331513269

Smart grid technologies can revolutionize power distribution. However, engineers can't write the software to realize this goal without large, publicly available datasets. Power companies are reluctant to share the data they collect because of consumer privacy concerns and concerns that they will expose themselves as targets of cyber attacks. We introduce a generative model for one of the most common types of smart grid data: synchrophasor network traffic. Our model can generate an unlimited amount of smart grid data while mitigating risks associated with exposing consumer and network information. Engineers can use our model to prototype smart grid software such as anomaly detection and state estimation. © 2025 IEEE.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Analyze and Construction of a Photovoltaic (PV) Power Plant Connected to the Grid to Produce Maximum Power on the Roof of the Building 9

Analyze and Construction of a Photovoltaic (PV) Power Plant ...

引用

9th International Conference on Technology and Energy Management, ICTEM 2024

作者： Karimi, Hamed Siadatan, Alireza Sepehrinour, Maryam Chaharmahali, Ehsan Shahid Beheshti University Department of Electrical and Computer Science Tehran Iran York University Department of Electrical and Computer Science Toronto Canada Western Kentucky University Department of Electrical Engineering Bowling GreenKY United States Science and Culture University Department of Electrical Engineering Tehran Iran

ISBN: (纸本)9798350329797

This paper is for technical studies and optimal selection for the angle of the panels and the feasibility of building a 20 kW power plant connected to the grid in the electricity distribution company of Tehran province with geographic coordinates of 1181 m above sea level, longitude and latitude 35° 41' 29'', 51° 26' 53'',taking into account the lack of shadow conditions and selecting the back-to-back distance in the rows of each panel for ease of maintenance and repairs and absorbing the most sunlight and producing it by photovoltaic arrays into energy and injecting it to the national grid. Also, the correct selection of electrical equipment such as inverter, panel, the optimal selection of the voltage produced by the strings compared to the working voltage of the inverter, which is one of the most important parts of the construction of this power *** power plant produces 35.920 MWh/yr of electricity annually and has reduced greenhouse gas emissions by 3875 Kg per month. Therefore, it can be effective in encouraging investment in the construction of these small power plants in the residential sector. and its results and the amount of production of this power plant during one year show the optimal design and construction of this solar site. © 2024 IEEE.

关键词： Solar power generation

来源：评论

学校读者我要写书评

暂无评论

Advancing Sustainable IoT Appliance Load Monitoring Through Edge-Enabled Federated Transfer Learning

Advancing Sustainable IoT Appliance Load Monitoring Through ...

引用

2024 International Conference on Green Energy, Computing and Sustainable Technology, GECOST 2024

作者： Natarajan, Yuvaraj Wadhwa, Gitanjali Sri Preethaa, K.R. Sundaram, Karthic Rama Abirami, K. Kpr Institute of Engineering and Technology Computer Science and Engineering Coimbatore641407 India Dresden01062 Germany Curtin University Faculty of Engineering and Science Department of Electrical and Computer Engineering Malaysia

ISBN: (纸本)9798350357905

Non-intrusive Appliance Load Monitoring (NALM) is essential for efficient electricity consumption tracking in households, promoting eco-friendly practices, and cost reduction. Privacy concerns in real-world NALM implementations can be addressed using federated learning. However, challenges persist, such as limited training data, edge model customization, and resource constraints. We introduce a practical federated learning framework for NALM, leveraging cloud-based model compression, collaborative data collection, and personalized edge and multi-task learning models through unsupervised transfer learning. Our advanced load monitoring model offers precise energy disaggregation through real-world experimentation, making it a leading solution for non-intrusive appliance load monitoring on edge devices. It maintains excellent learning performance and user privacy. Future research should aim to enhance federated learning efficiency and tackle remaining challenges in practical NALM implementations © 2024 IEEE.

关键词： Cost reduction

来源：评论

学校读者我要写书评

暂无评论

Optimizing Skin Lesion Segmentation with UNet and Attention-Guidance Utilizing Test Time Augmentation 6

Optimizing Skin Lesion Segmentation with UNet and Attention-...

引用

6th International Conference on electrical engineering and Information and Communication Technology, ICEEICT 2024

作者： Ahamed, Md. Faysal Hossain, Md. Munawar Mary, Mekhala Mariam Islam, Md. Rabiul Rajshahi University of Engineering & Technology Department of Electrical & Computer Engineering Rajshahi6204 Bangladesh International Standard University Department of Computer Science and Engineering Dhaka1212 Bangladesh

ISBN: (纸本)9798350385779

To properly diagnose and screen for skin disorders, precise segmentation of the affected area is required. Segmenting skin lesions is difficult because of factors such as the form of the lesion, its proximity to surrounding skin, and the presence of hair. To effectively tackle this matter, we present a pioneering technique that utilizes deep learning to achieve comprehensive automation in skin lesion segmentation. This approach incorporates advanced preprocessing methodologies to enhance the overall performance. The architecture Attention-UNet is employed for the purpose of selectively segmenting images of skin lesions. The high detection rate of this design allows for precise melanoma segmentation in real-time. Finally, in the post-processing phase, we applied test-time augmentation (TTA) to further enhance segmentation precision. Both the ISIC 2017 and ISIC 2018 skin lesion datasets were used for training and testing the suggested technique. The model's mDSC, mIoU, recall, and precision were all 91.93%, 88.15%, 92.72%, and 91.15%, respectively, when trained independently on the ISIC 2017 datasets. Separate training on the ISIC 2018 datasets achieved similar results: 92.17% mDSC, 90.26% mIoU, 92.34% recall, and 92.01% precision. After being trained on a dataset consisting ISIC data from both 2017 and 2018, the proposed method achieved mDSC scores of 92.30% percent, mIoU scores of 90.06% percent, recall rates of 93.91% percent, and precision rates of 90.74% percent. Our model demonstrates superior FPS scores compared to other models, indicating effective real-time performance. Due to its scalability and reliability, the suggested methodological framework may be utilized to create a completely automated computer-aided skin lesion diagnostic system. © 2024 IEEE.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Visually-Aware Context Modeling for News Image Captioning

Visually-Aware Context Modeling for News Image Captioning

引用

2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL 2024

作者： Qu, Tingyu Tuytelaars, Tinne Moens, Marie-Francine Department of Computer Science KU Leuven Belgium Department of Electrical Engineering KU Leuven Belgium

ISBN: (纸本)9798891761148

News Image Captioning aims to create captions from news articles and images, emphasizing the connection between textual context and visual elements. Recognizing the significance of human faces in news images and the face-name co-occurrence pattern in existing datasets, we propose a face-naming module for learning better name embeddings. Apart from names, which can be directly linked to an image area (faces), news image captions mostly contain context information that can only be found in the article. We design a retrieval strategy using CLIP to retrieve sentences that are semantically close to the image, mimicking human thought process of linking articles to images. Furthermore, to tackle the problem of the imbalanced proportion of article context and image context in captions, we introduce a simple yet effective method Contrasting with Language Model backbone (CoLaM) to the training pipeline. We conduct extensive experiments to demonstrate the efficacy of our framework. We outperform the previous state-of-the-art (without external data) by 7.97/5.80 CIDEr scores on GoodNews/NYTimes800k. ©2024 Association for Computational Linguistics.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Dynamic Model Selection for Asynchronous Federated Learning in IoBT Scenarios

Dynamic Model Selection for Asynchronous Federated Learning ...

引用

2024 IEEE International Conference on Big Data, BigData 2024

作者： Tankard, Earl Haileselassie Hagos, Desta Rawat, Danda B. Howard University Department of Electrical Engineering and Computer Science WashingtonDC United States

ISBN: (纸本)9798350362480

This paper explores the potential of Dynamic Model Selection (DMS) combined with asynchronous Federated Learning (FL) to enhance military devices within the Internet of Battlefield Things (IoBT). While previous studies have primarily focused on synchronous FL, we argue this approach has limitations. Hence, this study addresses this research gap by investigating asynchronous FL and its advantages over synchronous methods. Our main contribution lies in developing a dynamic client selection algorithm that optimizes the utility function, incorporating communication delay or latency, available computational resources, and model accuracy. This algorithm ensures clients are selected based on their potential contribution to the global model's performance, thereby improving the efficiency and effectiveness of the training process. Through experimentation with standard datasets (MNIST, CIFAR-10) and our custom dataset (Common Objects in Battlefield (COBA)) tailored for battlefield scenarios, we demonstrate the effectiveness of asynchronous FL and dynamic client selection. By addressing this critical research gap and providing empirical evidence, our study advances the use of FL for battlefield decision-making, significantly contributes to developing IoBT, and offers practical insights for military applications. © 2024 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Latent Variable and Classification Performance Analysis of Bird-Drone Spectrograms with Elementary Autoencoder

IEEE Transactions on Radar Systems

引用

IEEE Transactions on Radar Systems 2024年 3卷 115-123页

作者： White, Daniel Jahangir, Mohammed Mishra, Amit Kumar Baker, Chris J. Antoniou, Michail University Of Birmingham School Of Electrical Electronic And Systems Engineering United Kingdom Aberystwyth University Department Of Computer Science United Kingdom

Deep learning with convolutional neural networks has been widely utilised in radar research concerning automatic target recognition. Maximising numerical metrics to gauge the performance of such algorithms does not necessarily correspond to model robustness against untested targets, nor does it lead to improved model interpretability. Approaches designed to explain the mechanisms behind the operation of a classifier on radar data are proliferating, but bring with them a significant computational and analysis overhead. This work uses an elementary unsupervised convolutional autoencoder (CAE) to learn a compressed representation of a challenging dataset of urban bird and drone targets and subsequently if apparent quality of the representation via preservation of class labels leads to better classification performance after a separate supervised training stage. It is shown that a CAE that reduces the features output after each layer of the encoder gives rise to the best drone vs bird classifier. A clear connection between unsupervised evaluation via label preservation in the latent space and subsequent classification accuracy after supervised fine-tuning is shown, supporting further efforts to optimise radar data latent representations to enable optimal performance and model interpretability. © 2023 IEEE.

关键词： Drones

来源：评论

学校读者我要写书评

暂无评论

Very-Short-Term Solar Power Prediction Using a Suboptimal Multiple Fading Kalman Filter

Very-Short-Term Solar Power Prediction Using a Suboptimal Mu...

引用

2025 IEEE Texas Power and Energy Conference, TPEC 2025

作者： Nazaralizadeh, Solmaz Banerjee, Paramarshi Karimi, Sepehr Srivastava, Anurag K. Famouri, Parviz West Virginia University Lane Department of Computer Science and Electrical Engineering Morgantown United States

ISBN: (纸本)9798331541125

The uncertain nature and variability in solar power generation requires the development of methods to further improve the solar forecasting, specifically in very short term to drive the system operational decisions. With regard to this issue, this paper provides an innovative approach for predicting solar power in very short term using a Suboptimal Multiple Fading Extended Kalman Filter (SMFEKF). In this method, the fading factors are adjusted dynamically resulting in the SMFEKF having reduced computational burden as compared to traditional Kalman Filters such as Uncented Kalman Filter (UKF) and Extended Kalman Filter (EKF). Improving estimation precision is achieved through dynamically updating the covariance of process noise and measurement noise. In the proposed method, the historical solar irradiance data, numerical weather predictions, and the real-time power measurements are used to optimize prediction precision. The proposed SMFEKF demonstrates slightly better accuracy but significantly reduced computational burden and time. © 2025 IEEE.

关键词： Extended Kalman filters

来源：评论

学校读者我要写书评

暂无评论

Development of a String Transfer Assist Device for Electric Bass Picking 13

Development of a String Transfer Assist Device for Electric ...

引用

13th IEEE Global Conference on Consumer Electronic, GCCE 2024

作者： Marutsuka, Ryo Tsurumi, Kazuki Koutaki, Gou Kumamoto University Department of Computer Science and Electrical Engineering Kumamoto860-8555 Japan

ISBN: (纸本)9798350355079

This laboratory is developing a semi-automatic performance robot that involves human interaction. The robot is designed to provide an active performance experience. It is not fully automated, but rather semi-automated to allow for human involvement. While there have been robots in the past that assist with string pressing, playing the bass guitar can be challenging for beginners due to the need to pluck the strings separately with single notes. This presentation discusses the development of a device that facilitates picking motions in electric bass playing. © 2024 IEEE.

关键词： User centered design

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：