检索结果-内蒙古大学图书馆

European Wireless 2023;28th European Wireless conference

作者： Johannes Hofer Jannek Steinke Peter Sossalla Giang T. Nguyen Frank H. P. Fitzek

Edge-assisted visual Simultaneous Localization and Mapping (v-SLAM) systems offload complex analysis modules from computationally constrained mobile devices to cloud servers at the network edge. As visual data must travel through networks to reach the edge cloud, it is crucial to minimize transferred data while maintaining reliable and accurate localization and map generation. State-of-the-art systems perform pre-analysis of SLAM modules on the mobile device, causing high processing delays and limiting functionality. To minimise both throughput demand and latency, we are investigating video compression techniques in Compress-Then-Analyze (CTA) transmission systems. Evaluations employing the widely used H.264 video compression standard demonstrate that CTA achieves lower throughputs than its counterparts, as well as significantly reduced processing delays and transmission latencies. The results open up new possibilities to explore this research direction on edge-assisted v-SLAM further.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Electrical Performance analysis of Old Photovoltaic Modules under Outdoor Condition in Dhaka: Thermal Image-Based Investigation

Electrical Performance Analysis of Old Photovoltaic Modules ...

引用

International conference on Electrical, Computer and Communication Engineering (ECCE)

作者： Md. Sabbir Alam Ahmed Al Mansur Shahariar Ahmed Himo Md. Ashraful Islam Md. Mohi Uddin Mohin M M Naushad Ali Department of Electrical and Electronic Engineering Green University of Bangladesh Dhaka Bangladesh

ISBN: (数字)9798350357509

ISBN: (纸本)9798350357516

This study looks at the performance deterioration of six 10-watt photovoltaic (Pv) modules after five years in Bangladesh's humid environment. The research utilized a comprehensive approach, doing visual, thermal image, and electrical examinations to provide an integrated comprehension of the Pv module in outdoor conditions. A visual analysis of the top and rear views revealed various deterioration causes, including discoloration, delamination, back sheet damage, corrosion, and dust and hotspots. In the thermal image analysis of the Pv modules, the maximum temperature of the module was 60.70C, whereas the minimum was 28.30C. Simultaneously, electrical performance was assessed outside using an I-v tracer, and the experimental findings showed that the modules degraded over time. The highest output power for the old Pv panel was 9.584 watts, and the minimum was 5.766 watts. The highest yearly deterioration rate among the six panels was 8.468%, while the lowest was 0.832%. Over five years, the largest percentage deterioration rate was 42.34%, while the least was 4.16%. We measured the extent of performance reduction by comparing the collected data to the manufacturer's standards. This paper stresses the association between visual abnormalities and electrical degeneration, shining light on the interaction of these elements.

关键词： Degradation visualization Corrosion Thermal engineering Thermal conductivity Thermal analysis Solar panels Stress Meteorology Thermal degradation

来源：评论

学校读者我要写书评

暂无评论

Proceedings - 2021 IEEE visualization conference - Short Papers, vIS 2021

Proceedings - 2021 IEEE Visualization Conference - Short Pap...

引用

2021 IEEE visualization conference, vIS 2021

ISBN: (纸本)9781665433358

The proceedings contain 41 papers. The topics discussed include: understanding the effects of visualizing missing values on visual data exploration;screen parsing: towards reverse engineering of UI models from screenshots;an empirical evaluation of the T-SNE algorithm for data visualization in structural engineering;KNN learning techniques for proportional myocontrol in prosthetics;segmentation driven peeling for visual analysis of electronic transitions;hybrid data augmentation and deep attention-based dilated convolutional-recurrent neural networks for speech emotion recognition;and wellbeing supportive design - research-based guidelines for supporting psychological wellbeing in user experience.

关键词：

来源：评论

学校读者我要写书评

暂无评论

visual exploration of rating datasets and user groups

引用

FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE 2020年 105卷 547-561页

作者： Zegarra, Fabian Colque Carbajal Ipenza, Juan C. Omidvar-Tehrani, Behrooz Moreira, viviane P. Amer-Yahia, Sihem Comba, Joao L. D. Univ Fed Rio Grande do Sul Porto Alegre RS Brazil NAVER LABS Europe Meylan France Univ Grenoble Alpes Grenoble France CNRS Paris France

The increasing availability of rating datasets (i.e., datasets containing user evaluations on items such as products and services) constitutes a new opportunity in various applications ranging from behavioral analytics to recommendations. In this paper, we describe the design of vUGA, a visual enabler for the exploration of rating data and user groups. vUGA helps analysts, be they novice analysts or domain experts, acquire an understanding of their data through a seamless integration between exploring users and exploring their collective behavior via group analysis. vUGA is data-driven and does not require analysts to know the value distributions in their data. While automated systems can identify and suggest potentially interesting groups, they can do that for well-specified needs (e.g., through SQL QUERIES or constrained mining). vUGA helps analysts filter and refine their exploration as they discover what lies in the data. vUGA enables analysts to easily acquire statistics about their data, form groups, and find similar and dissimilar groups. While most visual analytics systems are data-dependent, vUGA relies on a data model that captures user data in such a way that a variety of group formation and exploration approaches can be used. We describe the architecture of vUGA and illustrate its use via tasks and a user study. We conclude with a discussion on future work enabled by vUGA. (C) 2019 Elsevier B.v. All rights reserved.

关键词： User data exploration User group exploration visual analytics User data Dimensionality reduction

来源：评论

学校读者我要写书评

暂无评论

Multimodal Apparent Personality analysis based on ConvNeXt-GRU and Random Forest

Multimodal Apparent Personality Analysis based on ConvNeXt-G...

引用

Machine Learning and Intelligent Systems Engineering (MLISE), International conference on

作者： Xiaodong Duan Jiaqi Shen Yonghui Yuan Feng Yang Yuangang Wang SEAC Key Laboratory of Big Data Applied Technology Dalian Minzu University Dalian China

ISBN: (数字)9798350375077

ISBN: (纸本)9798350375084

With the integrative development of computer science and psychology, apparent personality analysis has gradually become a new research hotspot. In this study, we aim to extract key information from large-scale video data and build an end-to-end deep learning model to explore the latent personality traits of vloggers in short videos. To fully utilize the spatiotemporal information existing in audio and visual modalities of short videos, we propose a multimodal apparent personality analysis method that integrates neural network and random forest. In visual modality analysis, we improve the ConvNeXt network by introducing the ECA attention mechanism and propose a new network structure unit named ECblock in order to obtain features closely relating to personality traits. Considering the temporal characteristics of short video, we combine the improved ConvNeXt with gated recurrent units for making personality traits assessment. For the audio modality, we manually extract voice features using librosa tool and employ random forest model for feature prediction under the help of grid search. Finally, according to decision fusion strategy, the latent information learned from visual and audio modalities are fused to implement apparent personality analysis. The results of comparative experiment on ChaLearn First Impressions v2 dataset show that our proposed method achieves competitive performance in terms of prediction accuracy.

关键词： visualization Analytical models Psychology Predictive models Feature extraction Brain modeling Robustness

来源：评论

学校读者我要写书评

暂无评论

Hybrid Line-Based and Region-Based Interactive Set data visualization

Hybrid Line-Based and Region-Based Interactive Set Data Visu...

引用

CHI conference on Human Factors in Computing Systems

作者： Wang, Xiaohan Zhang, Chuyu Zhu, Yu Chen, Xueyi Shen, Liming Liu, Richen Qian, Rongtao Nanjing Normal Univ Sch Artificial Intelligence Sch Comp & Elect Informat Nanjing Peoples R China

ISBN: (纸本)9781450380959

A generic step in data analysis is to group data items into multiple sets based on specific attribute values. In this paper, we propose an interactive set-data exploration tool named Balloonvis, to make nested balloons as a visual metaphor to group set data over a timeline and visualize the set overlapping information while preserving the original layout. We employ a hybrid region-based and line-based scheme, which allows placing a representative image of each set data item at its region position and helps reduce visual clutter by line connection design. Energy optimization is exploited to compute the layouts of region-based set data items (balloons) and connected lines. The case study and the user study suggest the Balloonvis can visualize set data with more information. Besides, the proposed hierarchical scheme is more scalable on set data item size.

关键词： set visualization region-based visualization line-based visualization

来源：评论

学校读者我要写书评

暂无评论

Universal visual Decomposer: Long-Horizon Manipulation Made Easy

Universal Visual Decomposer: Long-Horizon Manipulation Made ...

引用

IEEE International conference on Robotics and Automation (ICRA)

作者： Zichen Zhang Yunshuang Li Osbert Bastani Abhishek Gupta Dinesh Jayaraman Yecheng Jason Ma Luca Weihs Allen Institute for AI University of Pennsylvania University of Washington

ISBN: (数字)9798350384574

ISBN: (纸本)9798350384581

Real-world robotic tasks stretch over extended horizons and encompass multiple stages. Learning long-horizon manipulation tasks, however, is a long-standing challenge, and demands decomposing the overarching task into several manageable subtasks to facilitate policy learning and generalization to unseen tasks. Prior task decomposition methods require task-specific knowledge, are computationally intensive, and cannot readily be applied to new tasks. To address these shortcomings, we propose Universal visual Decomposer (UvD), an off-the-shelf task decomposition method for visual long-horizon manipulation using pre-trained visual representations designed for robotic control. At a high level, UvD discovers subgoals by detecting phase shifts in the embedding space of the pre-trained representation. Operating purely on visual demonstrations without auxiliary information, UvD can effectively extract visual subgoals embedded in the videos, while incurring zero additional training cost on top of standard visuomotor policy training. Goal-conditioned policies learned with UvD-discovered subgoals exhibit significantly improved compositional generalization at test time to unseen tasks. Furthermore, UvD-discovered subgoals can be used to construct goal-based reward shaping that jump-starts temporally extended exploration for reinforcement learning. We extensively evaluate UvD on both simulation and real-world tasks, and in all cases, UvD substantially outperforms baselines across imitation and reinforcement learning settings on in-domain and out-of-domain task sequences alike, validating the clear advantage of automated visual task decomposition within the simple, compact UvD framework.

关键词： Training visualization Costs Reinforcement learning Aerospace electronics data mining Task analysis

来源：评论

学校读者我要写书评

暂无评论

Evaluation of 5G Edge and Cloud Computing for data Processing in visual Referencing of Mobile Robot Manipulators 56

Evaluation of 5G Edge and Cloud Computing for Data Processin...

引用

56th CIRP International conference on Manufacturing Systems, CIRP CMS 2023

作者： Boshoff, Marius Schuster, David Christ, Lukas Hesenius, Marc Gruhn, volker Kuhlenkötter, Bernd Production Systems Ruhr University Bochum Bochum Germany Institute for Software Engineering University of Duisburg-Essen Essen Germany

To fit the requirements of growing flexibility in production processes, intralogistics is increasingly automated, for example, by Mobile Robot Manipulators, which can transport material and perform assembly tasks. However, these transport systems have limited battery capacity and recharging involves occasional downtime. New approaches, such as Edge Computing, offer transport systems to offload intensive computational processes, e.g., for path planning or AI-based image analysis, and thus saving precious energy resources. However, since offloading such processes onto a server is tied to high security and transmission requirements, using 5G for communication might be an attractive solution. Therefore, an exemplary referencing use case of a Mobile Robot Manipulator is presented and the 5G based communication with two cloud servers and one edge server is examined. Our experimental results show that with a mean transmission time of 540 milliseconds Edge Computing facilitates both faster and more stable data transmission compared to Cloud Computing with a mean transmission time of 1,973 milliseconds. This has a significant impact on industrial processes whose efficiency and productivity depend on the quality of data transfer. © 2023 The Authors. Published by Elsevier B.v. This is an open access article under the CC BY-NC-ND license (https://***/licenses/by-nc-nd/4.0) Peer-review under responsibility of the scientific committee of the 56th CIRP International conference on Manufacturing Systems 2023.

关键词： 5G Cloud Communication Edge Mobile Robotics

来源：评论

学校读者我要写书评

暂无评论

A Systematic Approach on Attention Based Multimodal Sentiment analysis in Convolutional Neural Network using Deep Learning Techniques

A Systematic Approach on Attention Based Multimodal Sentimen...

引用

Applied Intelligence and Computing (AIC), IEEE World conference on

作者： vinitha v S.K. Manju Bargavi Department of Computer Science and IT Jain (Deemed-to-be) University Bengaluru India

ISBN: (数字)9798350384598

ISBN: (纸本)9798350384604

Sentiment analysis represents an emerging field of study in both natural language processing and computational intelligence with consumer evaluations of products and services are advancing in autonomy. Multimodal sentiment analysis aims to determine a user’s sentiment using multimodal data. It seeks to describe a novel approach to sentiment categorization that employs deep learning and natural language processing. The proposed framework for attention-based multimodal sentiment classification combines convolutional neural networks for image data with bidirectional encoder representations from transformers for textual data. The synergy between these two modalities enhances the model’s capability of capturing abundant contextual information and visual signals. The framework employs a convolutional neural network for image data that recovers hierarchical features while preserving spatial relationships. Simultaneously, the BERT model encodes textual information, enabling the model to comprehend semantic sequence. Combining these modalities requires an attention mechanism to learn and dynamically evaluate the importance of visual and textual signals during classification. Attention-based fusion enhances sentiment categorization performance by focusing on the most informative elements from both modalities. Extensive evaluations utilizing the publicly available CMU Multimodal Opinion Sentiment and Emotion Intensity dataset demonstrate the utility of the proposed framework with an accuracy of 95.87%. The experimental findings indicate that the convolutional neural network with the BERT method captures complex sentiment patterns across various domains more effectively.

关键词： Deep learning Sentiment analysis visualization Attention mechanisms Accuracy Bidirectional control Transformers Encoding data models Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Design and exploration of Real Time Monitoring System for Power Grid Operation Safety Based on YOLO Object Detection Algorithm 3

Design and Exploration of Real Time Monitoring System for Po...

引用

3rd International conference on Machine Learning and Big data Analytics for IoT Security and Privacy, SPIoT2023

作者： Wu, Jianhui Shangqiu Power Supply Company State Grid Henan Electric Power Company Shangqiu China

As a fundamental industry of the national economy, the development level of the power industry directly affects the national economic construction and people's lives. With the continuous acceleration of modernization, higher requirements have been put forward for the safe operation of the PG. Therefore, it is necessary to achieve the integration of remote real-time monitoring, management, and maintenance under the current situation, and combine modern information technology methods for research and analysis. To meet the system requirements and ensure the quality of power supply, a suitable and efficient job safety management system platform is designed for the practical situation of power grid (PG) operation safety, and is based on the YOLO (You Only Look Once) algorithm operating environment. This platform can be effectively applied in daily electricity management work to improve the production efficiency and economic benefits of PG enterprises. Afterwards, this article also conducted functional testing on the security real-time monitoring system, and the test results showed that the accuracy of the system reached over 80%;the recall rate was over 85%;the average accuracy was between 0.4 and 0.7;the system stability was between 90% and 97%;the scalability of the system was over 93%. © 2023 Elsevier B.v.. All rights reserved.

关键词： Economic and social effects

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：