检索结果-内蒙古大学图书馆

data-Driven-Based Terahertz Image Restoration

ieee SENSORS JOURNAL 2023年第11期23卷 12284-12294页

作者： Su, Zhipeng Zhang, Yixiong Zhou, Jianyang Shi, Jianghong Qi, Feng Xiamen Univ Sch Informat Natl Demonstrat Software Sch Xiamen 361005 Peoples R China Xiamen Univ Sch Elect Sci & Engn Natl Model Microelect Coll Xiamen 361005 Peoples R China Xiamen Univ Sch Elect Sci & Engn Xiamen 361005 Peoples R China Chinese Acad Sci Inst Robot & Intelligent Mfg Key Lab Optoelect Informat Proc Shenyang Inst Automat Shenyang 110016 Peoples R China Chinese Acad Sci Key Lab Image Understanding & Comp Vis Shenyang 110016 Peoples R China

Terahertz (THz) has been widely used in radar, remote sensing, atmospheric and environmental monitoring, real-time biological information extraction, and medical diagnosis. However, the long-wavelength property of THz leads to low imaging resolution. Due to the expensive nature of experimental equipment and the difficulty of obtaining THz sources, most scientific institutions have difficulty in obtaining large amounts of the THz dataset. In addition, it is difficult for convolution neural networks (CNNs) to recover low-resolution complex images because of the phase information contained in the images. In this article, we build a THz imaging system and propose a progressive multiscale augmented neural network (PMANN) to address this problem. PMANN integrates self-attention and channel attention for image contextual feature extraction by the proposed progressive network. Compared with the complex CNN (CCNN) in private cartoon datasets, PMANN improves the quantitative metrics on peak signal-to-noise ratio (PSNR) by about 18.6%. It also obtains better visual quality than CCNN. Moreover, we publicly synthesize the first grayscale image MinistTHz dataset. Compared with SwinTransformer, PMANN improves the PSNR by 28.2%. The experimental results on the THz dataset demonstrate that the proposed model achieves better performances than the state-of-the-art (SOTA) models. Our synthetic THz dataset is open at https://***/szpxmu/PMANN.

关键词： Sensors Feature extraction Deep learning Convolution visualization Terahertz wave imaging Superresolution Low-resolution images progressive multiscale augmented neural network (PMANN) public datasets state-of-the-art (SOTA) terahertz (THz)

来源：评论

学校读者我要写书评

暂无评论

TinyHD: Efficient Video Saliency Prediction with Heterogeneous Decoders using Hierarchical Maps Distillation 23

TinyHD: Efficient Video Saliency Prediction with Heterogeneo...

引用

23rd ieee/CVF Winter Conference on Applications of Computer Vision (WACV)

作者： Hu, Feiyan Palazzo, Simone Salanitri, Federica Proietto Bellitto, Giovanni Moradi, Morteza Spampinato, Concetto McGuinness, Kevin Dublin City Univ Insight SFI Res Ctr Data Analyt Dublin Ireland Univ Catania PeRCeiVe Lab Catania Italy

ISBN: (纸本)9781665493468

Video saliency prediction has recently attracted attention of the research community, as it is an upstream task for several practical applications. However, current solutions are particurly computationally demanding, especially due to the wide usage of spatio-temporal 3D convolutions. We observe that, while different model architectures achieve similar performance on benchmarks, visual variations between predicted saliency maps are still significant. Inspired by this intuition, we propose a lightweight model that employs multiple simple heterogeneous decoders and adopts several practical approaches to improve accuracy while keeping computational costs low, such as hierarchical multi-map knowledge distillation, multi-output saliency prediction, unlabeled auxiliary datasets and channel reduction with teacher assistant supervision. Our approach achieves saliency prediction accuracy on par or better than state-of-the-art methods on DFH1K, UCF-Sports and Hollywood2 benchmarks, while enhancing significantly the efficiency of the model.

关键词： visualization Three-dimensional displays Computational modeling Computer architecture Predictive models Benchmark testing Semisupervised learning

来源：评论

学校读者我要写书评

暂无评论

Few-Shot Temporal Sentence Grounding via Memory-Guided Semantic Learning

引用

ieee TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY 2023年第5期33卷 2491-2505页

作者： Liu, Daizong Zhou, Pan Xu, Zichuan Wang, Haozhao Li, Ruixuan Huazhong Univ Sci & Technol Hubei Engn Res Ctr Big Data Secur Sch Cyber Sci & Engn Key Lab Distributed Syst Secur Wuhan 430074 Peoples R China Peking Univ Wangxuan Inst Comp Technol Beijing 100871 Peoples R China Dalian Univ Technol Sch Software Serv Software Liaoning Prov Key Lab Ubiquitous Network Dalian 116621 Peoples R China Huazhong Univ Sci & Technol Sch Comp Sci & Technol Wuhan 430074 Peoples R China

Temporal sentence grounding (TSG) is an important yet challenging task in video-based information retrieval. Given an untrimmed video input, it requires the machine to predict the interested video segment semantically related to a given sentence query. Most existing TSG methods train well-designed deep networks to align the semantic between video-query pairs for activity grounding with a large amount of data. However, we argue that these works easily capture the selection biases of video-query pairs in a dataset rather than showing the robust reasoning abilities to handle the rarely appeared pairs (i.e., few-shot contents). To alleviate such limitation of the off-balance data distribution during the network training, in this paper, we propose a novel memory-augmented network called Memory-Guided Semantic Learning Network (MGSL-Net) to handle the few-shot TSG task for enhancing the model generalization ability. Specifically, given the matched video-query input, we first employ a graph attentive cross-modal interaction module to align their semantics in a cycle-consistent manner. Then, we develop the memory modules in both video and query domains to record the cross-modal shared semantic features in the domain-specific persistent memory. At last, a heterogeneous attention module is utilized to integrate the memory-enhanced multi-modal features in both video and query domains with further feature calibration. During training, the memory modules are dynamically associated with both common and rare cases to memorize all appeared contents, alleviating the issue of forgetting the few-shot contents. Therefore, in testing, the rare cases can be enhanced by retrieving the stored memories, improving the generalization ability of the model. Experimental results on three benchmarks (ActivityNet Caption, Charades-STA and TACoS) show the superiority of our method on both effectiveness and efficiency.

关键词： Semantics Grounding Task analysis Training visualization Proposals Logic gates Temporal sentence grounding few-shot contents memory-augmented network

来源：评论

学校读者我要写书评

暂无评论

Research and Implementation of Movie Recommendation System Based on Knowledge Graph 4

Research and Implementation of Movie Recommendation System B...

引用

4th International Conference on Computer, Big data and Artificial Intelligence, ICCBD+AI 2023

作者： Luo, Lixia Huang, Zuoliu Tang, Qitao School of Computer Science and Engineering University of Information Technology Hunan Changsha China

ISBN: (纸本)9798350373233

Traditional recommendation algorithms based on the user's historical behavior and popularity of the product, which generally have the problems of data sparsity and cold start. Knowledge graph technology can provide users with richer and personalized recommendation results by integrating multidimensional features of movies and analyzing the associations between movies. This paper presents a recommendation algorithm which is based on knowledge graph technology, and analyses the requirements on the personalized movie recommendation system. And then based on the study of above, this paper designs the new whole system and then described the system implementation in detail. This recommendation system main functions include: Movies Management, Movies Details, Guess what you like, and Knowledge Graph visualization. Integrating knowledge graph technology into personalized movie recommendation system can effectively improve the accuracy and interpretability of movie recommendation. © 2023 ieee.

关键词： Recommender systems

来源：评论

学校读者我要写书评

暂无评论

High Performance EDA and LDA Analysis: An Application for Wheat Yield Estimation 3

High Performance EDA and LDA Analysis: An Application for Wh...

引用

3rd International Conference on Advances in Computing, Communication, Embedded and Secure Systems, ACCESS 2023

作者： Kumar, Deepak Kumar, Yash Kukreja, Vinay Bansal, Ankit Bhattacherjee, Abhishek Chitkara University Chitkara University Institute of Engineering and Technology Punjab India Lovely Professional University Department of Computer Science and Engineering Punjab Phagwara India

ISBN: (纸本)9798350346169

A worldwide industry that provides food, business, and employment opportunities, agriculture is a key component of human life. Despite this, wheat is one of the most common armed crops and the production rate harms wheat yield every year. In this paper, a prediction method for wheat yield has been calculated with different environmental impact assessment parameters. Predictors of data are a predictive approach that helps to categorize the data based on the different grouping patterns. Exploratory data analysis (EDA) and Linear discriminant analysis (LDA) are very effective approaches for grouping the data. The main aim of this paper is to predict the wheat yield prediction through EDA, decision tree, random forest regressor, ensemble learning, and LDA to maximize accuracy. Different environmental impacts parameters such as average rainfall, average temperature, and pesticides have been used to predict the wheat yield. Also, ensemble learning has been used for the prediction and analysis of the model through the decision tree and random forest regressor. Moreover, the LDA has been used to classify the wheat yield dataset by applying a reduction approach of LDA. During wheat yield prediction, the decision tree achieves 0.025 losses in training time. Also, the performance of LDA and EDA has been calculated through squared error functions. During wheat yield prediction through EDA with environmental impact parameters, the Root means squared error (RMSE) is 18245.27 while the value of Mean absolute error (MAE) is 12334.75. Furthermore, the work of LDA has presented by supporting the data visualization through different graphs using pandas and Matplotlib library. This study provides the data reduction predictors approach to the wheat yield and explains the data-preprocessing technique used along with EDA and LDA for wheat yield prediction in different environmental impact parameters. © 2023 ieee.

关键词： Crops

来源：评论

学校读者我要写书评

暂无评论

A Comprehensive Approach to Aquatic Environment Monitoring: IoT-based Smart Aquarium System

A Comprehensive Approach to Aquatic Environment Monitoring: ...

引用

2023 International Conference on Advances in Computation, Communication and Information Technology, ICAICCIT 2023

作者： Ghadekar, Premanand Khare, Sarthak Sakharwade, Dhawal Saraf, Chinmay Sonkamble, Pranav Amune, Amruta Vit Dept. of Information Technology Pune India

ISBN: (纸本)9798350344387

Aquaculture comprises a complex ecosystem involving aquatic organisms and their caregivers. Effective management of aquaculture significantly influences the well-being of aquatic life. With the continuous advancement of IoT (Internet of Things), AquaVision emerges as a potential catalyst for transforming aquarium management, establishing new benchmarks for both aquatic health and caretaker experiences. This research paper explores the conception, benefits, and limitations of AquaVision. The paper subsequently outlines the approach adopted for the creation and deployment of AquaVision, succeeded by an assessment of its merits and drawbacks. Current methodologies and trends in this domain are overviewed, while persisting challenges demanding resolution for successful adoption are underscored. Through the outcomes of the testing phase, AquaVision's efficacy as a tool for monitoring and controlling crucial aquarium parameters, encompassing pH, temperature, and water level, is substantiated. The system's successful integration and examination highlight its real-time monitoring competencies and remote-control functionalities. The discoveries from this study accentuate the expansive potential of IoT solutions in upholding public health and ensuring the accessibility of potable water. By elucidating AquaVision's evolution and benefits, the aspiration is to ignite fresh investigation and ingenuity within the realm of aquatic science and technology, ultimately fostering deeper comprehension, preservation, and admiration for the submerged world. © 2023 ieee.

关键词： Aquarium data visualization IoT Mobile application sensors

来源：评论

学校读者我要写书评

暂无评论

Machine Learning Models for Life Expectancy

Machine Learning Models for Life Expectancy

引用

2023 International Conference on Artificial Intelligence and Applications, ICAIA 2023 and Alliance Technology Conference, ATCON-1 2023

作者： Jalan, Deepanshi Tuli, Anandita Chaudhary, Vanshika Sharma, Nonita Rakhra, Manik Indira Gandhi Dtu for Women Department of Electronics and Communication with Ai Delhi India Indira Gandhi Dtu for Women Department of Information Technology Delhi India Lovely Professional University Department of Computer Science and Engineering Phagwara India

ISBN: (纸本)9781665456272

Life expectancy (LE) models provide a lot of ways to improve healthcare and other social welfares related to society. Life expectancy models provide solutions to problems like how to decide on retirement age or manage financial issues related to public matters. These models are becoming prominent in many regions as they are being widely used by government bodies and private sector for their policy making and developing health integrated systems. Thus, this paper aims to analyze the Trends in Life Expectancy in about 72 countries of the world over a span of 16 years, i.e., from 2000-2015. The study gives plots of attributes such as life expectancy, GDP, infant deaths, adult mortality, etc. across year which would help the countries understand the life expectancy trends over the course of time and suggest areas which should be focused upon to efficiently increase the life expectancy of its population. The simulations are done in Google Collab by using various Python libraries like pandas, numpy, matplotlib (used for plotting graphs), seaborn (used for plotting 3-D graphs and advanced visualization features of python), sklearn (used for handling missing data), and plotly express (used for plotting choropleth). © 2023 ieee.

关键词： Python

来源：评论

学校读者我要写书评

暂无评论

Beyond The Internet of Things: Imagining IoT's Structure and Potential Routes

Beyond The Internet of Things: Imagining IoT's Structure and...

引用

2023 International Conference on Advances in Computation, Communication and Information Technology, ICAICCIT 2023

作者： Varalakshmi, K.R. Mahesh, T.R. Singh, Akhilendra Pratap Presidency University Department of Electronics and Communication Engineering Karnataka Bangalore India Faculty of Engineering and Technology Department of Computer Science Engineering Karnataka Bangalore India Maharishi University of Information Technology Maharishi School of Engineering & Technology Uttar Pradesh India

ISBN: (纸本)9798350344387

wireless sensor system technology has made it possible for ubiquitous sensing, which affects many aspects of contemporary life. Indicators of environmental health, from sensitive ecologies as well as biodiversity to urban settings, may now be measured, inferred, and understood. The Internet of Things is enabled by the extensive utilization of these devices in a network that facilitates communication and action. In this network, the actuators and sensors smoothly interact with our environment as data is exchanged across platforms to provide a unified operational view. Near field communication (NFC) is the upcoming technology that will enable the transformation of the worldwide web into a fully integrated Future Internet. This technology is being advanced through the adoption of variousenabling device methods, such as tags, audience members, encased sense and actuators nodes, and nodes for near field interaction. The demand for data-on-demand, using complex and straightforward queries, increases significantly as studies transition from web to web2 to web3. This study presents a cloud-centric vision for the worldwideimplementation of the Internet of Things. The future of IoT research is expected to be influenced by many significant technology enablers and their corresponding application domains, which are emphasized. A cloud architecture based on the interplay of public as well as private clouds and employing Aneka is shown. Study conducted elaborate on the need of unity, the global web, and distributed processing in our last section of our IoT vision for the community of technical researchers. © 2023 ieee.

关键词： Cloud-centric Internet of Things data storage and analytics Internet of Things (IoT) New protocols Participatory Sensing Quality of Service (QoS) Radio Frequency Identification (RFID) User-cantered cloud-based paradigm visualization Wireless Sensor Networks

来源：评论

学校读者我要写书评

暂无评论

Ingest and Visualize CSV Files using AWS Platform for Transition from Unstructured to Structured data 11

Ingest and Visualize CSV Files using AWS Platform for Transi...

引用

11th International Conference on Emerging Trends in Engineering and Technology - Signal and Information Processing, ICETET - SIP 2023

作者： Gupta, Ayush Dhanda, Namrata Gupta, Kapil Kumar Amity University Computer Science & Engineering Department Uttar Prades Lucknow India

ISBN: (纸本)9798350348422

In addition to computing, storage, databases, analytics, networking, cellular, developer equipment, management equipment, IoT, security, and agency packages, AWS provides a comprehensive range of pay-as-you-go, on-demand, global cloud-based services. Accessible AWS services and solutions number over 200 and include information warehousing, directories, and content distribution. New offers can be immediately provisioned without an upfront capital cost. This makes it possible for businesses, start-ups, small and medium-sized organizations, and clients in the public sector to obtain access to the fundamentals needed to quickly adapt to changing business requirements. The technology at the core of any cloud operation is virtualization. It asserts that virtualization breaks up a physical server's hardware resources into smaller bits. data processing and visualization on Amazon Web Services is the process of gathering, analyzing, and visualizing large amounts of data using a variety of AWS services (AWS). data storage, integration, analysis, and visualization are just a few of the many services that Amazon offers that may be used to manage and display data. With the help of AWS services like Amazon S3, Amazon Kinesis, and Amazon Redshift, huge amounts of data can be acquired, analyzed, and stored. Amazon S3 is a simple storage service that may be used to store and retrieve data, Amazon Kinesis is a real-time data streaming service, and Amazon Redshift is a data warehousing service. Using AWS Glue, Redshift, and QuickSight, we demonstrate a method for processing and displaying CSV files in this paper In author's suggested approach, we extract, convert, and load data from CSV files to Redshift, a cloud-based data warehouse, using AWS Glue. The data saved in Redshift is then visualized using QuickSight, enabling quick and simple data exploration. We also go over how using AWS Glue, Redshift, and QuickSight can scale effectively and fast, automate data processing activities,

关键词： Digital storage

来源：评论

学校读者我要写书评

暂无评论

IFNet: An Interactive Frequency Convolutional Neural Network for Enhancing Motor Imagery Decoding From EEG

引用

ieee TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING 2023年 31卷 1900-1911页

作者： Wang, Jiaheng Yao, Lin Wang, Yueming Zhejiang Univ MOE Frontiers Sci Ctr Brain & Brain Machine Integr Hangzhou 310027 Peoples R China Zhejiang Univ Coll Comp Sci Hangzhou 310027 Peoples R China Zhejiang Univ Sch Med Affiliated Mental Hlth Ctr Dept Neurobiol Hangzhou 310030 Peoples R China Zhejiang Univ Hangzhou Peoples Hosp 7 Sch Med Hangzhou 310030 Peoples R China Zhejiang Univ QAAS Coll Comp Sci Hangzhou 310027 Peoples R China

Objective: The key principle of motor imagery (MI) decoding for electroencephalogram (EEG)-based Brain-Computer Interface (BCI) is to extract task-discriminative features from spectral, spatial, and temporal domains jointly and efficiently, whereas limited, noisy, and non-stationary EEG samples challenge the advanced design of decoding algorithms. Methods: Inspired by the concept of cross-frequency coupling and its correlation with different behavioral tasks, this paper proposes a lightweight Interactive Frequency Convolutional Neural Network (IFNet) to explore cross-frequency interactions for enhancing representation of MI characteristics. IFNet first extracts spectro-spatial features in low and high-frequency bands, respectively. Then the interplay between the two bands is learned using an element-wise addition operation followed by temporal average pooling. Combined with repeated trial augmentation as a regularizer, IFNet yields spectro-spatio-temporally robust features for the final MI classification. We conduct extensive experiments on two benchmark datasets: the BCI competition IV 2a (BCIC-IV-2a) dataset and the OpenBMI dataset. Results: Compared with state-of-the-art MI decoding algorithms, IFNet achieves significantly superior classification performance on both datasets while improving the winner's result in BCIC-IV-2a by 11%. Moreover, by conducting sensitivity analysis on decision windows, we show IFNet attains the best trade-off between decoding speed and accuracy. Detailed analysis and visualization verify IFNet can capture the coupling across frequency bands along with the known MI signatures. Conclusion: We demonstrate the effectiveness and superiority of the proposed IFNet for MI decoding. Significance: This study suggests IFNet holds promise for rapid response and accurate control in MI-BCI applications.

关键词： Decoding Electroencephalography Task analysis Brain modeling Feature extraction Convolution Convolutional neural networks Brain-computer interface motor imagery cross-frequency interactions convolutional neural networks data augmentation

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：