检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Appiani, Andrea Beyan, Cigdem Dep. of Management Information and Production Engineering University of Bergamo Bergamo Italy Department of Computer Science University of Verona Verona Italy

Voice Activity Detection (VAD) is the process of automatically determining whether a person is speaking and identifying the timing of their speech in an audiovisual data. Traditionally, this task has been tackled by processing either audio signals or visual data, or by combining both modalities through fusion or joint learning. In our study, drawing inspiration from recent advancements in visual-language models, we introduce a novel approach leveraging Contrastive Language-Image Pretraining (CLIP) models. The CLIP visual encoder analyzes video segments composed of the upper body of an individual, while the text encoder handles textual descriptions automatically generated through prompt engineering. Subsequently, embeddings from these encoders are fused through a deep neural network to perform VAD. Our experimental analysis across three VAD benchmarks showcases the superior performance of our method compared to existing visual VAD approaches. Notably, our approach outperforms several audio-visual methods despite its simplicity, and without requiring pre-training on extensive audio-visual datasets. © 2024, CC BY-NC-ND.

关键词： Visual languages

来源：评论

学校读者我要写书评

暂无评论

Investigating 3D reconstruction of non-collaborative surfaces through photogrammetry and photometric stereo 24

Investigating 3D reconstruction of non-collaborative surface...

引用

2021 24th ISPRS Congress Commission II: Imaging Today, Foreseeing Tomorrow

作者： Karami, A. Menna, F. Remondino, F. Trento Italy Dep. of Information Engineering and Computer Science University of Trento Trento Italy

3D digital reconstruction techniques are extensively used for quality control purposes. Among them, photogrammetry and photometric stereo methods have been for a long time used with success in several application fields. However, generating highly-detailed and reliable micro-measurements of non-collaborative surfaces is still an open issue. In these cases, photogrammetry can provide accurate low-frequency 3D information, whereas it struggles to extract reliable high-frequency details. Conversely, photometric stereo can recover a very detailed surface topography, although global surface deformation is often present. In this paper, we present the preliminary results of an ongoing project aiming to combine photogrammetry and photometric stereo in a synergetic fusion of the two techniques. Particularly, hereafter, we introduce the main concept design behind an image acquisition system we developed to capture images from different positions and under different lighting conditions as required by photogrammetry and photometric stereo techniques. We show the benefit of such a combination through some experimental tests. The experiments showed that the proposed method recovers the surface topography at the same high-resolution achievable with photometric stereo while preserving the photogrammetric accuracy. Furthermore, we exploit light directionality and multiple light sources to improve the quality of dense image matching in poorly textured surfaces. © 2021 International Archives of the Photogrammetry, Remote Sensing and Spatial information sciences - ISPRS Archives. All rights reserved.

关键词： Topography

来源：评论

学校读者我要写书评

暂无评论

A Latent Causal Inference Framework for Ordinal Variables

arXiv

引用

arXiv 2025年

作者： Scauda, Martina Kuipers, Jack Moffa, Giusi Dep. of Mathematics and Computer Science University of Basel Basel Switzerland Dep. of Pure Mathematics and Mathematical Statistics University of Cambridge Cambridge United Kingdom Dep. of Biosystems Science and Engineering ETH Zurich Basel Switzerland

Ordinal variables, such as on the Likert scale, are common in applied research. Yet, existing methods for causal inference tend to target nominal or continuous data. When applied to ordinal data, this fails to account for the inherent ordering or imposes well-defined relative magnitudes. Hence, there is a need for specialised methods to compute interventional effects between ordinal variables while accounting for their ordinality. One potential framework is to presume a latent Gaussian Directed Acyclic Graph (DAG) model: that the ordinal variables originate from marginally discretizing a set of Gaussian variables whose latent covariance matrix is constrained to satisfy the conditional indep.ndencies inherent in a DAG. Conditioned on a given latent covariance matrix and discretisation thresholds, we derive a closed-form function for ordinal causal effects in terms of interventional distributions in the latent space. Our causal estimation combines naturally with algorithms to learn the latent DAG and its parameters, like the Ordinal Structural EM algorithm. Simulations demonstrate the applicability of the proposed approach in estimating ordinal causal effects both for known and unknown structures of the latent graph. As an illustration of a real-world use case, the method is applied to survey data of 408 patients from a study on the functional relationships between symptoms of obsessive-compulsive disorder and dep.ession. © 2025, CC BY-NC-SA.

关键词： Directed graphs

来源：评论

学校读者我要写书评

暂无评论

RF-Based Lightweight Machine Learning for Comprehensive Drone Activity Classification

RF-Based Lightweight Machine Learning for Comprehensive Dron...

引用

Artificial Intelligence, Blockchain, and Internet of Things (AIBThings), IEEE International Conference on

作者： Mohammed A. Alqodah Mahdia Tahsin Mohammad H. Omari Mustafa M. Matalgah David Harrison Dep. of Electrical & Computer Engineering University of Mississippi University MS USA Dept. of Computer and Information Science University of Mississippi University MS USA

ISBN: (数字)9798331529529

ISBN: (纸本)9798331529536

The rapid rise of commercially available drones, driven by miniaturization and affordability, has introduced potential security risks alongside numerous societal benefits. While drones serve various industries, their unregulated use necessitates robust detection and classification methods. This paper proposes a novel, efficient model for classifying unwanted drone activity. We leverage statistical and time-domain features extracted from radio frequency (RF) signals and machine learning (ML) classifiers. This approach aims to not only detect drones but also distinguish their models and operational modes. Utilizing the DroneRF dataset, our feature selection strategy prioritizes informative signals in the low-band spectrum, achieving ${100\%}$ accuracy in both drone detection and drone model identification tasks. Furthermore, our hybrid model, combining time-domain and frequency-domain features, achieves an accuracy of ${97.66\%}$ in classifying different flight modes. This demonstrates the effectiveness of this comprehensive feature extraction approach for drone activity classification.

关键词： Radio frequency Accuracy RF signals Neural networks Feature extraction Real-time systems Security Time-domain analysis Random forests Drones

来源：评论

学校读者我要写书评

暂无评论

Using the web as corpus for un-supervised learning in question answering 16

Using the web as corpus for un-supervised learning in questi...

引用

16th Conference on Computational Linguistics and Speech Processing, ROCLING 2004

作者： Wang, Yi-Chia Wu, Jian-Cheng Liang, Tyne Chang, Jason S. Dep. of Computer and Information Science National Chiao Tung University Taiwan Dep. of Computer Science National Tsing Hua University Taiwan

In this paper we propose a method for unsupervised learning of relation between terms in questions and answer passages by using the Web as corpus. The method involves automatic acquisition of relevant answer passages from the Web for a set of questions and answers, as well as alignment of wh -phrases and keywords in questions with phrases in the answer passages. At run time, wh-phrases and keywords are transformed to a sequence of expanded query terms in order to bias the underlying search engine to give higher rank to relevant passages. Evaluation on a set of questions shows that our prototype improves the performance of a question answering system by increasing the precision rate of top ranking passages returned by the search engine. © Proc. of the 16th Conference on Computational Linguistics and Speech Processing, ROCLING 2004. All rights reserved.

关键词： Search engines

来源：评论

学校读者我要写书评

暂无评论

Implementation and Design of Physical/Digital Switches Modified from Sonoff Product Based on Internet of Things 4

Implementation and Design of Physical/Digital Switches Modif...

引用

4th IEEE Eurasia Conference on Biomedical engineering, Healthcare and Sustainability, ECBIOS 2022

作者： Tsao, Yung-Chung Kuo, Yaw-Wen Wu, Chia-Chun Tsai, Yin-Te Hsu, Chihcheng Li, Chiyang Wang, Han-Ching National Chi Nan University Dep. of Electrical Engineering Nantou County Taiwan National Quemoy University Dep. of Industrial Engineering and Management Kinmen County Taiwan Providence University Dep. of Computer Science and Communication Engineering Taichung City Taiwan National Central University Dep. of Information Management Taoyuan Country Taiwan College of Science and Technology National Chi Nan University Nantou County Taiwan Xinhe Elementary School Xindian District New Taipei City Taiwan

ISBN: (数字)9781728195797

ISBN: (纸本)9781728195797

While more home appliances are designed and manufactured, Sonoff serial home assistant alliances are designed and manufactured by eWeLink for many cheap and convenient devices controlled with exclusive APPs. Sonoff home assistant alliances are popular and well-known to attract many developers who can develop unofficial software to enhance the closed APPs designed by eWeLink and make flexible applications based on original hardware such as Tasmota and parameterized open-source APPs. We follow the above-mentioned developers to design a firmware that controls and is controlled based on message queue telemetry transport (MQTT) protocol from a cloud platform and the internet of thing (IoT) based architecture. The Sonoff home assistant alliances installed by the designed firmware can be controlled with one or more groups through a clouding platform by existing physical power-sockets from one-to-one controlling by eWeLink APPs. Under the proposed architecture, those Sonoff serial home assistant alliances are easily used like traditional power sockets for old-age users, who only use physical switches with touch-panel with MCU-based controllers to control power sockets. We designed digital switches on the website by browsers and physical touch panels as traditional switches work indep.ndently and coexist to meet most users with software-based APPs. The proposed architecture can be implemented and dep.oyed as a proof of concept (PoC) for the usability and validity. © 2022 IEEE.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

A User Experience Study on University of Isfahan's LMS and Exploring Features of AI-Enhanced E-Learning Platforms

A User Experience Study on University of Isfahan's LMS and E...

引用

International Conference on E-Learning and E-Teaching (ICELET)

作者： Milad Ghaffari Ali Akbari Morteza Mohammadi Ostani Dep. of Mechatronics Engineering Islamic Azad University Majlesi Branch Isfahan Iran Dep. of Knowledge & Information Science University of Isfahan Isfahan Iran

The main objective of this paper was to evaluate the user experience (UX) of students in higher education by conducting usability testing in the e-learning context, specifically utilizing the University of Isfahan (UI) Learning Management System (LMS). The survey method was used in the current study and to collect data we utilized random sampling to evaluate the usability aspects defined in ISO 9241-11 standards based on a modified system usability scale (SUS) involving 377 end-users. The data were analyzed using Microsoft Excel software and findings showed that SUS score for the University of Isfahan's LMS is 60.05 which falls in the “OK usability” range and is acceptable. Also, we explored the features of AI-enhanced LMSs using a ranking question to comprehensively understand students' perspectives suggesting that Chatbots for support and intelligent content recommendations are among the most expected features for prominent educational outcomes. From the NPS assessment, -43 implies areas for improvement.

关键词：

来源：评论

学校读者我要写书评

暂无评论

IoT-Enabled Low-Cost Fog Computing System with Online Machine Learning for Accurate and Low-Latency Heart Monitoring in Rural Healthcare Settings

arXiv

引用

arXiv 2023年

作者： Maneshti, Hamidreza Dadashi, Morteza Rostami, Kamyar Dept. Computer Eng. & Information Tech Engineering Kermanshah Iran Dep. Computer Science Montreal Canada

Healthcare services in rural areas face numerous challenges due to the high cost of treatment and a lack of appropriate services. The application of Internet of Things (IoT) technology has shown potential in mitigating these issues. This article discusses the potential of Internet of Things (IoT) and fog computing to reduce healthcare costs and improve patient outcomes. The use of these technologies in cardiovascular health informatics is explored, along with the economic thought process of hospital decision-makers and end-of-life practices in intensive care units. Remote monitoring using IoT devices is highlighted as a promising way to detect health issues before they become serious, leading to earlier interventions and improved health outcomes. The use of fog computing in healthcare is also discussed, with a focus on its ability to provide real-time data processing, analysis, and decision-making capabilities. The article presents a novel architecture for Device-as-a-Service, utilizing both fog and cloud computing to improve the efficiency and accuracy of ECG device processing, and concludes that it has the potential to reduce costs by up to 80% in the Iranian market. The adoption of fog computing in healthcare is acknowledged to present significant challenges, such as security and privacy concerns, but its potential benefits make it a promising technology for the future of healthcare services. © 2023, CC BY.

关键词： Electrocardiograms

来源：评论

学校读者我要写书评

暂无评论

Review of Cloud AI for Real-Time Fraud Detection 10

Review of Cloud AI for Real-Time Fraud Detection

引用

10th International Conference on Automation, Robotics, and Applications, ICARA 2024

作者： Aabitova, Gulnara Abalkanov, Miras Abitova, Gulnara A. Shuteyeva, Gulnur Aitmukhanbetova, Elvira Kulniyazova, Korlan Astana IT University Dep. Computer Engineeering Astana Kazakhstan State Univerity of New York at Binghamton Dept. Electical Engineering and Computer Science Binghamton United States Eurasian National University Dept. System Anslysis and Managemnt Astana Kazakhstan

ISBN: (纸本)9798350394245

The paper presents a comprehensive analysis of the integration of Cloud AI in real-time fraud detection, highlighting significant advancements and practical applications. It is identifying a key challenge such as data privacy and the need for scalable solutions, while also exploring the potential of innovative AI algorithms and strategic partnerships to enhance fraud detection capabilities. Through detailed case studies, the work is demonstrated the effective application of Cloud AI across various sectors. The study concludes that Cloud AI, despite not being a complete solution, plays a crucial role in strengthening fraud detection systems. Our findings suggest that strategic implementation of Cloud AI can lead to more adaptive and efficient systems, thereby significantly contributing to the economic sector's resilience against fraudulent activities. This paper underscores the transformative impact of Cloud AI in fraud detection, offering insights into its future potential and laying the groundwork for further advancements in this field. © 2024 IEEE.

关键词： Crime

来源：评论

学校读者我要写书评

暂无评论

Implementation and Design of Water Flow Meter of Smart Home Integrated with Environment Monitor System 4

Implementation and Design of Water Flow Meter of Smart Home ...

引用

4th IEEE Eurasia Conference on Biomedical engineering, Healthcare and Sustainability, ECBIOS 2022

作者： Tsao, Yung-Chung Kuo, Yaw-Wen Wu, Chia-Chun Tsai, Yin-Te Hsu, Chihcheng Ho, Liulin Wang, Han-Ching National Chi Nan University Dep. of Electrical Engineering Nantou County Taiwan National Quemoy University Dep. of Industrial Engineering and Management Kinmen County Taiwan Providence University Dep. of Computer Science and Communication Engineering Taichung City Taiwan National Central University Dep. of Information Management Taoyuan Country Taiwan College of Science and Technology National Chi Nan University Nantou County Taiwan Xinhe Elementary School Xindian District New Taipei City Taiwan

ISBN: (数字)9781728195797

ISBN: (纸本)9781728195797

More home appliances were designed and manufactured in the market, but fewer commercial products related to water-flow monitoring are sold in the current market. Water-related costs and expenses are cheaper than others, but water is one of the critical survival necessities for human beings. Thus, the home appliances or the smart home needs to pay attention to ubiquitous water resources as one of the limited natural earth resources. We follow the above-mentioned issues to achieve an internet of things (IoT) based architecture to design a flow-monitoring information collecting device (FMICD). Hall sensors were used to collect the water-flow consumption and to upload the information of the environment monitor system which was implemented in the previous research. In the proposed architecture, the FMICDs are designed and developed with the MCU-based controller binding to a water-flow meter (Model: YF-S201) equipped to monitor water consumption. We implemented FMICDs in the Xinhe Elementary School in the Xindian District of New Taipei City in Taiwan to validate the usability and feasibility of the study with a website (http://***:9999/***) to offer related users and staff to validate the usability in the study. © 2022 IEEE.

关键词： Internet of things

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：