检索结果-内蒙古大学图书馆

International Conference on Power Energy, Environment and Intelligent Control (PEEIC)

作者： Deepti Mounika Nalluri Chinna babu Mupparaju Aruna Sri Rongali Sohong Dhar P. Ajitha Sloan School of Management Sloan Fellow Candidate 2023 MIT Masters in Computer Information Systems Murray State University USA Masters in Computer Science university of central Missouri warrensburg USA Graduate Doctoral Information Technology/Graduate (Information Technology Emphasis/Information Technology Ph.D.) University of The Cumberlands US Azure Microsoft library & information science Jadavpur University India School of Computing Sathyabama Institute of science and technology Chennai

ISBN: (数字)9798350357769

ISBN: (纸本)9798350357776

Deep learning is a powerful subset of synthetic intelligence, and it has the potential to significantly enhance the skills of machine getting to know. it's miles primarily based on the idea of artificial neural networks, which might be laptop fashions that are capable of gaining knowledge of from records and making decisions, just like the manner a human mind works. Deep mastering isn't the same as traditional system mastering algorithms in that it can technique complex inputs with multiple layers of abstraction, main to greater accurate evaluation. Deep studying algorithms are capable of unexpectedly examine raw facts and generate styles, making them perfect for actual-time statistics analysis. Deep studying is capable of as it should be predicting behaviors and trends, which may be used to optimize procedures and enhance operational efficiencies. Deep studying models can also be used for automated fault detection and predictive renovation, permitting businesses to count on and prevent capacity troubles. Additionally, deep mastering models frequently offer quicker consequences than other device studying or conventional techniques, as well as permitting an extra diploma of control over the resolution of the data being analyzed. This has widespread implications for real-time programs, such as medical diagnostics, independent automobiles, and manufacturing monitoring.

关键词：

来源：评论

学校读者我要写书评

暂无评论

An Analysis Method for Metric-Level Switching in Beat Tracking

arXiv

引用

arXiv 2022年

作者： Chiu, Ching-Yu Müller, Meinard Davies, Matthew E.P. Su, Alvin Wen-Yu Yang, Yi-Hsuan Graduate Program of Multimedia Systems and Intelligent Computing National Cheng Kung University Academia Sinica Taiwan International Audio Laboratories Erlangen Germany Department of Informatics Engineering Centre for Informatics and Systems The University of Coimbra University of Coimbra Portugal Department of CSIE National Cheng Kung University Taiwan Taiwan AI Labs Taiwan

For expressive music, the tempo may change over time, posing challenges to tracking the beats by an automatic model. The model may first tap to the correct tempo, but then may fail to adapt to a tempo change, or switch between several incorrect but perceptually plausible ones (e.g., half- or double-tempo). Existing evaluation metrics for beat tracking do not reflect such behaviors, as they typically assume a fixed relationship between the reference beats and estimated beats. In this paper, we propose a new performance analysis method, called annotation coverage ratio (ACR), that accounts for a variety of possible metric-level switching behaviors of beat trackers. The idea is to derive sequences of modified reference beats of all metrical levels for every two consecutive reference beats, and compare every sequence of modified reference beats to the subsequences of estimated beats. We show via experiments on three datasets of different genres the usefulness of ACR when being utilized alongside existing metrics, and discuss the new insights that can be gained. © 2022, CC BY.

关键词： Taps

来源：评论

学校读者我要写书评

暂无评论

Analyzing the Impact of Deep Learning Approaches on Real-Time Data Analysis in Machine Learning

Analyzing the Impact of Deep Learning Approaches on Real-Tim...

引用

International Conference on Power Energy, Environment and Intelligent Control (PEEIC)

关键词：

来源：评论

学校读者我要写书评

暂无评论

High-Precision Stereo Matching with a Drone Equipped With Rtk-Gnss and a Single Camera for Plant Height Measurements

SSRN

引用

SSRN 2022年

作者： Matsuura, Yuta Heming, Zhang Nakao, Kousuke Qiong, Chang Firmansyah, Iman Kawai, Shin Yamaguchi, Yoshiki Hayashi, Hisayoshi Maruyama, Tsutomu Nobuhara, Hajime The Department of Intelligent and Mechanical Interaction Systems Graduate School of Science and Technology University of Tsukuba 1-1-1 Tennoudai Ibaraki Tsukuba305-8573 Japan College of Engineering Systems School of Science and Engineering University of Tsukuba Japan The Department of Computer Science Graduate School of Science and Technology University of Tsukuba Japan The Faculty of Life and Environmental Sciences University of Tsukuba Japan The School of Computing Tokyo Institute of Technology Japan

We propose a high-precision plant height measurement system based on large-scale stereo images captured by a drone. In our system, the drone collects images of a field using a monocular camera and obtains the position data by a real time kinematic (RTK)-global navigation satellite system (GNSS). The heights of plants in the field are measured by stereo matching. By employing a single monocular camera rather than a binocular camera, we can obtain a sufficient baseline length for accurate height measurements, but the images must be precisely calibrated at run time. To this end, we propose a calibration algorithm with a conventional stereo matching algorithm. The proposed method requires only a pair of images and their position data (RTK-GNSS), which are more suitable for edge computing on drones than three-dimensional (3D) reconstructions by methods such as Structure from Motion. To show the effectiveness of the proposed method, we measured the plant height from photographs captured by the drone at different altitudes. Our method achieves a higher height resolution (1.6 cm) than the 3D reconstruction-based methods Pix4Dmapper and Metashape, with error-rate reductions of 62.2% and 69.4%, respectively, in the 10-20 m altitude range. When the altitude is lowered to 4.1 m , the depth resolution is improved to 1.6 mm with error reductions of 44.4% and 63.0% from those of Pix4Dmapper and Metashape, respectively. On a GTX 1080 TI graphics processing unit, stereo matching of 5472 × 3468-pixel images consumed 88 ms. This execution time is sufficiently fast for real-time measurements. Because our system uses the position data obtained by RTK-GNSS, it achieves centimeter-level accuracy during the drone flight. After the flight, the accuracy can be improved to millimeter level by comparing the results of the stereo matching with an image containing a reference plant of preobtained known height © 2022, The Authors. All rights reserved.

关键词： Global positioning system

来源：评论

学校读者我要写书评

暂无评论

Improvement of Spatial Ambiguity in Multi-Channel Speech Separation Using Channel Attention

Improvement of Spatial Ambiguity in Multi-Channel Speech Sep...

引用

Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA)

作者： Qian-Bei Hong Chung-Hsien Wu Thanh Binh Nguyen Hsin-Min Wang Graduate Program of Multimedia Systems and Intelligent Computing National Cheng Kung University and Academia Sinica Tainan Taiwan National Cheng Kung University Tainan Taiwan

ISBN: (纸本)9781665441629

Multi-channel speech separation has been successfully applied in a complex real-world environment such as the far-field condition. The common solution to deal with the far-field condition is using a multi-channel signal captured by a structured microphone array and leveraging the inner difference between channels to enhance the speech separation performance. The spatial feature has been widely used in recent speech separation research. This feature appears to be insufficient when the location information becomes ambiguous. This is known as the spatial ambiguity problem. In order to deal with the spatial ambiguity problem, this study proposes an attention mechanism for the Temporal-Spatial Neural Filter (TSNF), in which the channel attention on merged features and the feature map of 1D convolution block in the temporal convolution network is proposed. The proposed method is evaluated on the multi-channel reverberant dataset which is built based on the WSJ0-2mix dataset. The dataset is simulated in the real-environment room by using the Room Impulse Response generator. In the experimental results, the proposed methods produced the SI -SNR improvement of about 1.2dB in close speakers' case, while a small decrease of 0.1dB in other cases.

关键词： Convolution Information processing Speech enhancement Information filters Microphone arrays Generators Reliability

来源：评论

学校读者我要写书评

暂无评论

Communication and Control Co-Design in 6G: Sequential Decision-Making with LLMs

引用

IEEE Network 2024年

作者： Chen, Xianfu Wu, Celimuge Shen, Yi Ji, Yusheng Yoshinaga, Tsutomu Ni, Qiang Zarakovitis, Charilaos C. Zhang, Honggang Shenzhen CyberAray Network Technology Co. Ltd China Chinese Academy of Sciences Shanghai Advanced Research Institute China Meta-Networking Research Center China The University of Electro-Communications Graduate School of Informatics and Engineering Tokyo Japan National Institute of Informatics Information Systems Architecture Research Division Tokyo Japan Lancaster University School of Computing and Communications Ukraine AXON Logic Greece City University of Macau Faculty of Data Science Macau China China

This article investigates a control system within the context of sixth-generation wireless networks. The remote control performance optimization confronts the technical challenges that arise from the intricate interactions between communication and control sub-systems, asking for a co-design. Considering the system dynamics, we formulate the sequential co-design decision-makings of communication and control over a discrete time horizon as a Markov decision process, for which a practical offline learning framework is proposed. Our proposed framework integrates large language models into the elements of reinforcement learning. We present a case study on the age of semantics-aware communication and control co-design to showcase the potential of our proposed learning framework. Furthermore, we discuss the open issues remaining to make our offline learning framework feasible for real-world implementations and highlight the research directions for future explorations. © 1986-2012 IEEE.

关键词： Markov processes

来源：评论

学校读者我要写书评

暂无评论

High-Precision Stereo Matching with a Drone Equipped With Rtk-Gnss and a Single Camera for Plant Height Measurements

SSRN

引用

SSRN 2022年

关键词： Global positioning system

来源：评论

学校读者我要写书评

暂无评论

Covert Channel Detection and Generation Techniques: A Survey

Covert Channel Detection and Generation Techniques: A Survey

引用

Emerging Smart Technologies and Applications (eSmarTA), International Conference on

作者： Nesmah A. AL-Khulaidi Ammar T. Zahary Muneer A.S Hazaa Adel A. Nasser Department of Computer Networks Yemen Academy for Graduate Studies Sana'a Yemen Department of Computer Science Faculty of Computing and IT University of Science and Technology Sana'a Yemen Department of Computer Science Faculty of Computer and Information System Thamar University Thamar Yemen Department of Information Systems and Computer Science Faculty of Sciences Sa'adah University Sa'adah Yemen

The use of covert communications has become more widespread recently as a solution to the information security issue. Information security can be partially solved by the discovery and creation of covert routes. Covert channel use is still in its early phases, even though information security concerns are growing as a result of quick technology advancements. As a result, several problems and concerns need to be resolved, and the future plan is foggy. This paper seeks to give a thorough comparison of earlier and more recent attempts in this field. It offers a survey of the literature on the drawbacks and gaps in conventional detection techniques, emphasizing the most recent techniques for discovering and creating these covert channels while carefully considering machine learning and deep learning-based detection techniques. Additionally, it broadens the topic to cover the crucial function that these channels play in technology. Timing channels, storage channels, and hybrid channels are just a few of the ways covert channels may be utilized to transfer data. They can be used to get around security measures like intrusion detection systems and firewalls. Additionally, they may be used to steal confidential data from a system, including passwords, credit card numbers, and intellectual property. System security is seriously threatened by covert channels, and more study on covert channel production and detection is required to stay up with the changing threat environment.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Delivering Sensing Technologies for Education and Learning

Delivering Sensing Technologies for Education and Learning

引用

Adjunct Proceedings of the 2022 ACM International Joint Conference on Pervasive and Ubiquitous computing and the 2022 ACM International Symposium on Wearable Computers

作者： Andrew Vargo Victoria Abou-Khalil Shoya Ishimaru Benjamin Tag Mathilde Hutin Andreas Dengel Laurence Devillers Koichi Kise Graduate School of Informatics Osaka Metropolitan University Japan ETH Zurich Switzerland German Research Center for Artificial Intelligence (DFKI) Germany and University of Kaiserslautern Germany School of Computing and Information Systems University of Melbourne Australia LISN-CNRS France Smart Data & Knowledge Services German Research Center for Artificial Intelligence (DFKI) Germany and Knowledge Based Systems University of Kaiserslautern Germany Paris-Sorbonne University France and LISN-CNRS France

ISBN: (纸本)9781450394239

Pervasive and ubiquitous sensing technologies have been frequently leveraged in experimental settings within the Ubicomp, ISWC, and HCI community. In the recent past, we have seen advances in the use of ubiquitous sensing technologies to understand and support learning. Recent work investigated a wide range of education-related technologies that detect attention, postures, behaviors, and emotions. However, the promise of delivering these technologies to potential users en masse remains difficult to fulfill. At this time it is important to understand the factors hindering the use of these technologies, to critically evaluate methodologies for technological development, and to discuss in what ways developed technologies can be safely delivered to students. In this workshop, we bring together experts from the fields of ubiquitous and pervasive computing, HCI, and education to discuss and develop an agenda for moving sensing technologies for learning and education from research to practice.

关键词： augmented learning pervasive computing physiological sensing for education

来源：评论

学校读者我要写书评

暂无评论

Speaker-Specific Articulatory Feature Extraction Based on Knowledge Distillation for Speaker Recognition

引用

APSIPA Transactions on Signal and Information Processing 2023年第2期12卷

作者： Qian-Bei Hong Hsin-Min Wang Chung-Hsien Wu Graduate Program of Multimedia Systems and Intelligent Computing National Cheng Kung University and Academia Sinica Taiwan Graduate Program of Multimedia Systems and Intelligent Computing National Cheng Kung University and Academia Sinica and Department of Computer Science and Information Engineering National Cheng Kung University Taiwan

This paper proposes a novel speaker-specific articulatory feature (AF) extraction model based on knowledge distillation (KD) for speaker recognition. First, an AF extractor is trained as a teacher model for extracting the AF profiles of the input speaker dataset. Next, a KD-based speaker embedding extraction method is proposed to distill the speaker-specific information from the AF profiles in the teacher model to a student model based on multi-task learning, in which the lower layers not only capture the speaker characteristics from acoustic features, but also learn the speaker-specific features from the AF profiles for robust speaker representation. Finally, speaker embeddings are extracted from the high-level layer, and the obtained speaker embeddings are further used to train a probabilistic linear discriminant analysis (PLDA) model for speaker recognition. In the experiments, speaker embedding models were trained using the VoxCeleb2 dataset and the AF extractor was trained based on the LibriSpeech dataset, and the performance was evaluated using the VoxCeleb1 dataset. The experiments showed that the proposed KD-based models outperformed the baseline models without KD. Furthermore, feature concatenation of multimodal results can further improve the performance.

关键词： Speaker recognition articulatory feature knowledge distillation

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：