检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Wang, Chien-Chun Chen, Li-Wei Lee, Hung-Shin Chen, Berlin Wang, Hsin-Min Dept. Computer Science and Information Engineering National Taiwan Normal University Taiwan Institute of Computer Science Academia Sinica Taiwan United-Link Co. Ltd. Taiwan

Cross-domain speech enhancement (SE) is often faced with severe challenges due to the scarcity of noise and background information in an unseen target domain, leading to a mismatch between training and test conditions. This study puts forward a novel data simulation method to address this issue, leveraging noise-extractive techniques and generative adversarial networks (GANs) with only limited target noisy speech data. Notably, our method employs a noise encoder to extract noise embeddings from target-domain data. These embeddings aptly guide the generator to synthesize utterances acoustically fitted to the target domain while authentically preserving the phonetic content of the input clean speech. Furthermore, we introduce the notion of dynamic stochastic perturbation, which can inject controlled perturbations into the noise embeddings during inference, thereby enabling the model to generalize well to unseen noise conditions. Experiments on the VoiceBank-DEMAND benchmark dataset demonstrate that our domain-adaptive SE method outperforms an existing strong baseline based on data simulation. © 2024, CC BY.

关键词： Background noise

来源：评论

学校读者我要写书评

暂无评论

Improving the Accessibility of Online Training: A Usability Evaluation of NetAcad Training for Users with Disabilities

Improving the Accessibility of Online Training: A Usability ...

引用

International Conference on Computational science and Computational Intelligence (CSCI)

作者： Heba Ismail Ahmad Jasim Jasmy Mohammad Bilal Mohammad Rasool Dept. of Computer Science and Information Technology Abu Dhabi University Abu Dhabi United Arab Emirates

ISBN: (数字)9798350361513

ISBN: (纸本)9798350372304

This study evaluates the usability of NetAcad, a web-based engineering training platform, among users with disabilities. A usability questionnaire is distributed to 42 participants, including users with visual impairment, motor disabilities, ADHD, and normal users. The collected data was analyzed using quantitative methods to determine the violated usability heuristics in NetAcad from a user perspective. Users with visual impairments and motor disabilities found it most challenging to use the NetAcad platform. The study identified specific usability issues faced by users with disabilities when interacting with the platform, such as lack of instant visual feedback, low visibility of course progress during the online training session, and the lack of accessibility features supporting people with visual impairments. A one-sample t-test showed that users with disabilities (vision impairment, M=2.6; ADHD, M=2.5; motor disability, M=2.9) had a significantly lower user experience when compared to normal users (M=3.4) considering the target value (M=5). To improve the accessibility and usability of online training platforms, the study recommends several accessibility features that accommodate users with disabilities.

关键词： Training Surveys Visualization Scientific computing Visual impairment Motors User experience

来源：评论

学校读者我要写书评

暂无评论

SIT-HANDS: A Novel RGB-D Dataset for Static Hand Gesture Recognition Using Kinect Sensor

SIT-HANDS: A Novel RGB-D Dataset for Static Hand Gesture Rec...

引用

Smart Systems for applications in Electrical sciences (ICSSES), International Conference on

作者： Dayananda Kumar N. C K.V Suresh Sheela S. J Dinesh R Dept. of Electronics and Communication Engineering Siddaganga Institute of Technology Tumkur India Dept. of Information Science and Engineering Jain University Bangalore India

Hand gesture recognition refers to the classification of various hand poses which conveys certain encoded signals used non-verbal communication. Hand gestures are also used in verbal communication to emphasize the meaning of spoken words. Hand gesture recognition finds various applications like human computer interaction and sign language recognition. It is an active area of research on solving the real time challenges like varying lighting conditions and uncluttered environment. Currently most of the datasets used in vision based hand gesture recognition are captured using RGB camera in lab environment which does not simulate the real time use case. Hence, we propose a novel dataset SIT-HANDS captured using Kinect V2 to obtain RGB frame and corresponding dept. map data pair for static hand gesture recognition. Proposed dataset consists of 14 unique gestures captured from 10 different subjects with varying background and illumination.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Resource Allocation for OTFS-Based ISAC Systems

Resource Allocation for OTFS-Based ISAC Systems

引用

Communications and Networking (MeditCom), IEEE International Mediterranean Conference on

作者： Meiying Zhang Hengtao He Shenghui Song Jun Zhang Khaled B. Letaief Dept. of Electronic and Computer Engineering The Hong Kong University of Science and Technology Hong Kong

Integrated sensing and communications (ISAC) has been considered as a key technology for future wireless networks. Orthogonal time frequency space (OTFS) modulation is considered as an effective technology to achieve ISAC in high-mobility scenarios due to its ability to combat high delay and Doppler shifts. However, although the performance analysis of OTFS-based ISAC system has been investigated, resource allocation and joint optimization for communication and sensing are still in its fancy. In this paper, we consider resource allocation in OTFS-based ISAC systems. We first introduce the system model and formulate the resource allocation problems. Then, several related works are discussed in detail. Next, an OTFS-based multiple-user multiple-input multiple-output (MIMO) ISAC system model is investigated, where the base station is considered as an ISAC transmitter and sends the message to different vehicles while simultaneously estimating their relative range and velocity via output feedback, respectively. We then formulate the sensing and communication trade-off optimization problem considering the delay Doppler resource block (DDRB) allocation and provide some sample results. Finally, we show several future directions for resource allocation in OTFS-based ISAC systems.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Design and Implementation of a Distributed Network Management System for a Multi-Branch University using GNS3

Design and Implementation of a Distributed Network Managemen...

引用

International Conference on Computing and Networking technology (ICCNT)

作者： Md Apu Hosen Shahadat Hoshen Moz Md. Ali Akber Dept. of Computer Science and Engineering Northern University of Business and Technology Khulna Khulna Bangladesh

The rapid digitalization of communication necessitates continuous skill updates to keep pace with evolving technology. Network management involves multiple functionalities that play a crucial role in optimizing network performance and ensuring continuous availability. It is very important to ensure a proper network management system for a university that has multiple branches to effectively manage and monitor their network infrastructure. A proper network management system provides centralized control and visibility over the entire network infrastructure, allowing administrators to efficiently manage and monitor network devices, configurations, and performance from a single interface. this research has developed a network management system that facilitates real-time network monitoring and troubleshooting, enabling proactive identification and resolution of issues to ensure smooth operations across all branches. It simplifies configuration management, ensuring consistency and compliance in network settings and updates. Resource optimization is achieved through insights into network utilization and performance metrics, allowing administrators to allocate resources efficiently and plan for future expansion. Security and compliance measures can be enforced, protecting sensitive data and ensuring adherence to regulatory requirements. This study utilized GNS3 to create a network layout for a university with multiple branches, along with essential security configurations for the network design. A diverse range of protocols was implemented to safeguard and cater to the users of the secure university network. Furthermore, Switches, Data Routers, firewalls, and BGP Routers were incorporated into the network to enhance communication efficiency and streamline operations.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Yoga Pose Assessment for Self-Learning 1

Yoga Pose Assessment for Self-Learning

引用

1st International Conference on Smart and Sustainable Technologies in Energy and Power Sectors, SSTEPS 2022

作者： Anthoniraj, S. Kumar, A Naresh Reddy, Galiveeti Hemakumar Raju, More Dept. of Computer Science and Engineering Karnataka Bangalore India Institute of Aeronautical Engineering Dept. of Electrical and Electronics Engineering Hyderabad India Akhildev IPR and Research Services Department of Research and Development Greater Noida India Maulana Azad National Institute of Technology Dept. of Electrical Engineering Madhya Pradesh Bhopal India

ISBN: (纸本)9781665464147

Amid the coronavirus pandemic, more number of people started practicing yoga by themselves by watching online videos because going to classes is not feasible these days. This paper provides a review on yoga pose self-Assessment system, which assesses a yoga practitioner's pose correctness with the help of deep learning algorithms, so that injuries caused out of incorrect postures can be prevented. The system detects a yoga pose and then calculates the difference between the pose of the user and that of the instructor using dynamic time warping (DTW) for video input and using angle difference calculation for static image input. Our primary contributions include a novel solution that employs a lightweight pose estimation model and DTW for time series analysis of the videos. © 2022 IEEE.

关键词： Coronavirus

来源：评论

学校读者我要写书评

暂无评论

A Hybrid Federated Learning Model for Insurance Fraud Detection

A Hybrid Federated Learning Model for Insurance Fraud Detect...

引用

IEEE International Conference on Communications Workshops, ICC

作者： Supriya Y Nancy Victor Gautam Srivastava Thippa Reddy Gadekallu School of Information Technology and Engineering Vellore Institute of Technology Vellore Tamil Nadu India Dept. of Math and Computer Science Brandon University Brandon MB Canada

Mission-critical systems are significant for the survival of any organization. Financial systems, communication systems, and electricity grid systems are some examples of mission-critical applications. Any damage to such systems can cause business operations to fail. Insurance fraud is one of the challenges faced by mission-critical systems today. Insurance fraud detection has historically been a manual task left to claim agents, who look over the evidence and draw an inference based on their intuitions. Recently, machine learning (ML) has been predominantly used to automate the process of insurance fraud detection. Despite the success of ML models in detecting insurance fraud, privacy preservation of insured personal data continues to be a problem. To address this issue, we present a novel hybrid approach. The proposed work suggests an automated way to control the process of vehicle insurance claim fraud detection in the insurance industry. The proposed study combines Federated Learning (FL), the Genetic Algorithm (GA), and Particle Swarm Optimization (PSO) to incorporate the advantages of each technology. The suggested model uses GA for optimal feature subset extraction. The optimized feature subset is then fed into a Federated learning with a Particle swarm optimization (FPSO) model. The findings show that the suggested hybrid model has an accuracy of 94.47% and that it may be improved even further by using other nature-inspired algorithms that are only used for fraud detection.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Transpose-SELDNet for Polyphonic Sound Event Localization and Detection

A Transpose-SELDNet for Polyphonic Sound Event Localization ...

引用

International Conference for Convergence of technology (I2CT)

作者： Spoorthy V Shashidhar G. Koolagudi Dept. of Computer Science and Engineering National Institute of Technology Karnataka Surathkal Mangalore India

Human beings have the ability to identify a particular event occurring in a surrounding based on sound cues even when no visual scenes are presented. Sound events are the auditory cues that are present in a surrounding. Sound event detection (SED) is the process of determining the beginning and end of sound events as well as a textual label for the event. The term sound source localization (SSL) refers to the process of identifying the spatial location of a sound occurrence in addition to the SED. The integrated task of SED and SSL is known as Sound Event Localization and Detection (SELD). In this proposed work, three different deep learning architectures are explored to perform SELD. The three deep learning architectures are SELDNet, D-SELDNet (dept.wise Convolution), and T-SELDNet (Transpose Convolution). Two sets of features are used to perform SED and Direction-of-Arrival (DOA) estimation tasks in this work. D-SELDNet uses a dept.wise convolution layer which helps reduce the model’s complexity in terms of computation time. T-SELDNet uses Transpose Convolution, which helps in learning better discriminative features by retaining the input size and not losing necessary information from the input. The proposed method is evaluated on the First-order Ambisonic (FOA) array format of the TAU-NIGENS Spatial Sound Events 2020 dataset. An improvement has been observed as compared to the existing SELD systems with the proposed T-SELDNet.

关键词： Location awareness Visualization Direction-of-arrival estimation Convolution Event detection Deep architecture Estimation

来源：评论

学校读者我要写书评

暂无评论

A New Benchmark on Musculoskeletal Abnormality Recognition System using Deep Transfer Learning Model

A New Benchmark on Musculoskeletal Abnormality Recognition S...

引用

International Conference on Advancement in Electrical and Electronic engineering (ICAEEE)

作者： Tajmin Sultana Mime Diponkor Bala Mohammad Alamgir Hossain Md. Atiqur Rahman Md. Shamim Hossain Md. Ibrahim Abdullah Dept. of Computer Science and Engg. Islamic University Kushtia Bangladesh School of Information Science and Tech. University of Science and Technology of China Hefei Anhui China

ISBN: (数字)9798350388282

ISBN: (纸本)9798350388299

Musculoskeletal disorders are abnormalities of the bones, muscles, and joints that affect the great majority of people worldwide. Radiographic scans are the most widely used technique for detecting these aberrations as part of medical diagnostics. Early-stage detection of anomalies in radiographs is crucial for the patient. Moreover, bone fracture classification is costly, more time-consuming, and requires more effort. These reasons have made the deep learning-based classifier model a reliable alternative. In this study, we have applied a deep transfer learning technique to the problem of classifying such anomalies in radiographs of the musculoskeletal system. For multiclass recognition of radiological images into seven categories, the suggested model has been executed on the widely used MURA dataset, which includes 40,561 radiograph images. We utilized the DenseNet121 pre-trained model with some optimization in the suggested model. Our proposed model achieved the highest accuracy of $96.62 \%$. The new model achieved a precision of $96.73 \%$, a recall of $96.62 \%$, an $F 1$-score of $96.60 \%$, and a Cohen Kappa score of $96.06 \%$. We attain better performances that approximately $3 \%$ exceed the previous study and are also comparable to state-of-the-art performances.

关键词： Accuracy Image recognition Magnetic resonance imaging Computed tomography Transfer learning Bones Reliability

来源：评论

学校读者我要写书评

暂无评论

Feel Good AI: Voice-Enabled Emotion-based Music Recommendation System

Feel Good AI: Voice-Enabled Emotion-based Music Recommendati...

引用

Advances in Computing, Communication and Applied Informatics (ACCAI), International Conference on

作者： Anjali Sharma Shubham Vishwakarma Liya T Mathew Dept. of Computer Science and Engg JSS Academy of Technical Education Bangalore India Dept. of Civil and Engg Thakur College of Engineering and Technology Mumbai India Dept. of Electronics and Communication Engg Cochin University of Science and Technology Kochi India

ISBN: (数字)9798350389449

ISBN: (纸本)9798350389456

Feel Good AI is a novel effort committed to improving psychological well-being by means of methodically providing customized music suggestions that are in accordance with the user’s emotional state. By combining sophisticated recommendation algorithms with advanced technologies, most notably Convolutional Neural Networks (CNN) for facial emotion recognition, this project effortlessly develops innovative tech and empathetic emotional support. The users interact with an intuitive chatbot that generates unique music suggestions through the use of personalized conversations in which they are asked specific emotion-related inquiries. Beyond immediate relevance, incorporating AI-powered music curation to enhance positive emotional experiences for individuals, particularly those afflicted with dementia, is of greater significance. We live in a society where stress and mental health problems are prevalent. By combining problem-solving and emotional well-being, Feel Good AI offers evidence that technology can enhance the quality of life.

关键词： Emotion recognition Technological innovation Temperature Mood Refining Real-time systems Convolutional neural networks

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：