检索结果-内蒙古大学图书馆

Archiving 2021

作者： Deborah, Hilda Mandai, Dipendra J. Norwegian Colour and Visual Computing Laboratory Department of Computer Science NTNU Norwegian University of Science and Technology Gjovik Norway

ISBN: (纸本)9780892083541

Microfiche was a common format used in microforms reproductions of documents, extensively used for archival storage before the move to digital formats. While contemporary documents are still available for digitization, others from older historical periods are no longer physically accessible for various reasons. In some cases, their microfiche copies are available, making microfiche digitization a must. However, a microfiche reader is not always available and, even then, it is a machine made for the purpose of reading and not for data collection. In this work, the performance two imaging devices are evaluated as alternatives to the traditional microfiche reader, by means of optical character recognition (OCR). Results show that this alternative surpasses the performance of a microfiche reader in terms of text legibility. © 2021 Society for Imaging science and Technology.

关键词： Digital storage

来源：评论

学校读者我要写书评

暂无评论

New Problems in Active Sampling for Mobile Robotic Online Learning

New Problems in Active Sampling for Mobile Robotic Online Le...

引用

IEEE Annual International computer Software and Applications Conference (COMPSAC)

作者： Xiuxian Guan Junming Wang Zekai Sun Zongyuan Zhang Tianyang Duan Shengliang Deng Fangming Liu Heming Cui Department of Computer Science The University of Hong Kong Hong Kong China Peng Cheng Laboratory Shenzhen China

AI models deployed in real-world tasks (e.g., surveillance, implicit mapping, health care) typically need to be online trained for better modelling of the changing real-world environments and various online training methods (e.g., domain adaptation, few shot learning) are proposed for refining the AI models based on training input incrementally sampled from the real world. However, in the whole loop of AI model online training, there is a section rarely discussed: how to sample training input from the real world. In this paper, we show from the perspective of online training of AI models deployed on edge devices (e.g., robots) that several problems in sampling of training input on the device are affecting the time and energy consumption for the online training process to reach high performance. Notably, the online training relies on training input consecutively sampled from the real world and the consecutive samples from nearby states (e.g., position and orientation of a camera) are too similar and would limit the training accuracy gain per training iteration; on the other hand, while we can choose to sample more about the inaccurate samples to better final training accuracy, it is costly to obtain the accuracy statistics of samples via traditional ways such as validating, especially for AI models deployed on edge devices. These findings aim to raise research effort for practical online training of AI models, so that they can achieve resiliently and sustainably high performance in real-world tasks.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Enhanced Redundant Weighted Clustered Scheme for Damaged S-UAV

Enhanced Redundant Weighted Clustered Scheme for Damaged S-U...

引用

IEEE International Conference on Communications (ICC)

作者： Grace Khayat Constandinos X Mavromoustakis Andreas Pitsillides Jordi Mongay Batalla Evangelos K. Markakis Department of Computer Science Mobile Systems Laboratory (MoSys Lab) University of Nicosia Nicosia Cyprus Department of Computer Science Cyprus and Department of Electrical & Electronic Engineering Science University of Cyprus University of Johannesburg (Visiting Professor) Warsaw University of Technology Warsaw Poland Department of Electrical and Computer Engineering Hellenic Mediteranean University Estavromenos Heraklion Greece

A spontaneous group of unmanned aerial vehicles (UAVs) is denoted as a swarm of UAVs (S-UAVs). The UAVs communicate wirelessly and cooperate to accomplish tasks. In crisis scenarios such as flooding or earthquakes, all UAVs are at risk of getting damaged and thus non-functional. A non- functional UAV will result in a disconnected network, especially if that UAV is highly responsible for packet forwarding. S- UAV s are dynamic networks; clustering is one of the most adopted routing schemes in S-UAVs. The clustering scheme groups the UAVs into clusters where each cluster is formed of a cluster head (CH) and cluster members (CMs). Only the CH can handle inter-cluster communication. Due to the crucial role played by the CH, its selection is a continuous field of research. This paper proposes an enhanced clustered weighted scheme with redundancy to ensure end-to-end communication. The proposed scheme is based on a weighted formula for the primary CH, redundant CH, and CMs selection. The weighted formula calculates a cluster index based on the distance, the speed, and the reward index. A new component is added to the reward index which is performance. The redundant CH is selected to automatically replace the primary CH whenever it is damaged. If the redundant CH becomes inoperable, the second redundant CH will take over. Each cluster is formed of n CMs and will have n-2 redundant CHs. The results obtained from the conducted simulation experiments concluded that this promising scheme decreases data loss in a crisis case scenario.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Trends and Topics: Characterizing Echo Chambers' Topological Stability and In-group Attitudes

arXiv

引用

arXiv 2023年

作者： Cau, Erica Morini, Virginia Rossetti, Giulio Computer Science Department University of Pisa Italy KDD Laboratory ISTI National Research Council Pisa Italy

Social Network sites are fertile ground for several polluting phenomena affecting online and offline spaces. Among these phenomena are included echo chambers, closed systems in which the opinions expressed by the people inside are exacerbated for the effect of the repetition, while opposite views are actively excluded. This paper offers a framework to explore, in a platform-independent manner, the topological changes of echo chambers through time while considering the content posted by users and the attitude conveyed in discussing specific controversial issues. The proposed framework consists of four steps: (i) data collection and annotation of users' ideology regarding a controversial topic, (ii) construction of a dynamic network of interactions, (iii) echo chambers extraction and analysis of their dynamics, and (iv) topic extraction and valence analysis. The paper then enhances the formalization of the framework by conducting a case study on Reddit threads about sociopolitical issues (gun control, American politics, and minorities discrimination) during the first two years and a half of Donald Trump's presidency. The results unveil that users often stay inside echo chambers over time. Furthermore, the analyzed discussions focus on controversies related to right-wing parties and specific events in American and Canadian politics. The analysis of the attitude conveyed in the discussions shows a slight inclination toward a more negative or neutral attitude when discussing particularly sensitive issues, such as fascism, school shootings, or police violence. © 2023, CC BY.

关键词： Topology

来源：评论

学校读者我要写书评

暂无评论

Rethinking Image Forgery Detection via Contrastive Learning and Unsupervised Clustering

arXiv

引用

arXiv 2023年

作者： Wu, Haiwei Chen, Yiming Zhou, Jiantao State Key Laboratory of Internet of Things for Smart City Department of Computer and Information Science University of Macau China

Image forgery detection aims to detect and locate forged regions in an image. Most existing forgery detection algorithms formulate classification problems to classify pixels into forged or pristine. However, the definition of forged and pristine pixels is only relative within one single image, e.g., a forged region in image A is actually a pristine one in its source image B (splicing forgery). Such a relative definition has been severely overlooked by existing methods, which unnecessarily mix forged (pristine) regions across different images into the same category. To resolve this dilemma, we propose the FOrensic ContrAstive cLustering (FOCAL) method, a novel, simple yet very effective paradigm based on contrastive learning and unsupervised clustering for the image forgery detection. Specifically, FOCAL 1) utilizes pixel-level contrastive learning to supervise the high-level forensic feature extraction in an image-by-image manner, explicitly reflecting the above relative definition;2) employs an on-the-fly unsupervised clustering algorithm (instead of a trained one) to cluster the learned features into forged/pristine categories, further suppressing the cross-image influence from training data;and 3) allows to further boost the detection performance via simple feature-level concatenation without the need of retraining. Extensive experimental results over six public testing datasets demonstrate that our proposed FOCAL significantly outperforms the state-of-the-art competing algorithms by big margins: +24.3% on Coverage, +18.6% on Columbia, +17.5% on FF++, +14.2% on MISD, +13.5% on CASIA and +10.3% on NIST in terms of IoU. The paradigm of FOCAL could bring fresh insights and serve as a novel benchmark for the image forgery detection task. The code is available at https://***/HighwayWu/FOCAL. © 2023, CC BY-NC-ND.

关键词： Pixels

来源：评论

学校读者我要写书评

暂无评论

Learning Perceptual Hallucination for Multi-Robot Navigation in Narrow Hallways

Learning Perceptual Hallucination for Multi-Robot Navigation...

引用

IEEE International Conference on Robotics and Automation (ICRA)

作者： Jin-Soo Park Xuesu Xiao Garrett Warnell Harel Yedidsion Peter Stone Department of Electrical and Computer Engineering The University of Texas at Austin United States Department of Computer Science George Mason University United States Everyday Robots United States Department of Computer Science The University of Texas at Austin United States Army Research Laboratory United States Sony AI Japan

While current systems for autonomous robot navigation can produce safe and efficient motion plans in static environments, they usually generate suboptimal behaviors when multiple robots must navigate together in confined spaces. For example, when two robots meet each other in a narrow hallway, they may either turn around to find an alternative route or collide with each other. This paper presents a new approach to navigation that allows two robots to pass each other in a narrow hallway without colliding, stopping, or waiting. Our approach, Perceptual Hallucination for Hallway Passing (PHHP), learns to synthetically generate virtual obstacles (i.e., perceptual hallucination) to facilitate passing in narrow hallways by multiple robots that utilize otherwise standard autonomous navigation systems. Our experiments on physical robots in a variety of hallways show improved performance compared to multiple baselines.

关键词：

来源：评论

学校读者我要写书评

暂无评论

An Intelligent Transportation Systems-Based Machine Learning-Enhanced Traffic Prediction Model using Time Series Analysis and Regression Techniques*

An Intelligent Transportation Systems-Based Machine Learning...

引用

IEEE International Symposium on Innovations in Intelligent Systems and Applications (INISTA)

作者： Klea Elmazi Donald Elmazi Eugen Musta Fatjon Mehmeti Florenc Hidri Department of Computer Engineering Faculty of Engineering University of Rijeka University Luarasi Tirana Albania Department of Computer Engineering Faculty of Computer Science and IT Artificial Intelligence Laboratory University Metropolitan Tirana Tirana Albania Department of Finance Faculty of Economy Canadian Institute of Technology Tirana Albania Department of Computer Engineering University Metropolitan Tirana Tirana Albania Department of Software Engineering Faculty of Engineering Canadian Institute of Technology Tirana Albania

ISBN: (数字)9798350368130

ISBN: (纸本)9798350368147

Traffic congestion represents a daunting challenge for all facets of urban development, as well as represents a universal problem in all urban areas, to various extents. In recent years, many cities have adopted the use of Intelligent Transport Systems (ITS) to manage traffic congestion. These systems are indeed useful, but they are mainly geared to predict real-time traffic congestion, yielding to a certain short-sightedness in our prediction model. Through the use of Time Series Analysis, and Regression models for traffic congestion prediction, we posit that we can address the issue. The data that these models were trained on derives from two datasets, namely a dataset from Kaggle and another from the road traffic footage of Tirana. The data then points to Gated Recurrent Units (GRU) being a more accurate time series category, and Support Vector Regression (SVR) to be a better performer than linear regression.

关键词： Support vector machines Analytical models Technological innovation Time series analysis Urban areas Transportation Predictive models

来源：评论

学校读者我要写书评

暂无评论

Co-Evolution of Pose and Mesh for 3D Human Body Estimation from Video

Co-Evolution of Pose and Mesh for 3D Human Body Estimation f...

引用

International Conference on computer Vision (ICCV)

作者： Yingxuan You Hong Liu Ti Wang Wenhao Li Runwei Ding Xia Li Key Laboratory of Machine Perception Shenzhen Graduate School Peking University Department of Computer Science ETH Zurich

Despite significant progress in single image-based 3D human mesh recovery, accurately and smoothly recovering 3D human motion from a video remains challenging. Existing video-based methods generally recover human mesh by estimating the complex pose and shape parameters from coupled image features, whose high complexity and low representation ability often result in inconsistent pose motion and limited shape patterns. To alleviate this issue, we introduce 3D pose as the intermediary and propose a Pose and Mesh Co-Evolution network (PMCE) that decouples this task into two parts: 1) video-based 3D human pose estimation and 2) mesh vertices regression from the estimated 3D pose and temporal image feature. Specifically, we propose a two-stream encoder that estimates mid-frame 3D pose and extracts a temporal image feature from the input image sequence. In addition, we design a co-evolution decoder that performs pose and mesh interactions with the image-guided Adaptive Layer Normalization (AdaLN) to make pose and mesh fit the human body shape. Extensive experiments demonstrate that the proposed PMCE outperforms previous state-of-the-art methods in terms of both per-frame accuracy and temporal consistency on three benchmark datasets: 3DPW, Human3.6M, and MPI-INF-3DHP. Our code is available at https://***/kasvii/PMCE.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Security and Privacy Evaluation Methods and Metrics in Vehicular Networks 19

Security and Privacy Evaluation Methods and Metrics in Vehic...

引用

19th IEEE Annual Consumer Communications and Networking Conference, CCNC 2022

作者： Benarous, Leila Boudjit, Saadi Faculty of Science University of Amar Telidji Lim Laboratory Laboratoire d'Informatique et de Mathématique Department of Computer Science Laghouat Algeria University of Sorbonne Paris Nord L2TI Laboratory Villetaneuse93430 France

ISBN: (纸本)9781665431620

The vehicular networks extend the internet services to road edge. They allow users to stay connected offering them a set of safety and infotainment services like weather forecasts and road conditions. The security and privacy are essential issues in computing systems and networks. They are particularly important in vehicular networks due to their direct impact on the users' safety on road. Various researchers have concentrated their efforts on resolving these two issues in vehicular networks. A great number of researches are found in literature and with still existing open issues and security risks to be solved, the research is continuous in this area. However, the researchers may face some difficulties in choosing the correct method to prove their works or to illustrate their excellency in comparison with existing solutions. In this paper, we review a set of evaluation methodologies and metrics to measure, proof or analyze privacy and security solutions. The aim of this review is to illuminate the readers about the possible existing methods to help them choose the correct techniques to use and reduce their difficulties. © 2022 IEEE.

关键词： Weather forecasting

来源：评论

学校读者我要写书评

暂无评论

Cumulus cloud modeling from images based on VAE-GAN

引用

Virtual Reality & Intelligent Hardware 2021年第2期3卷 171-181页

作者： Zili ZHANG Yunchi CEN Fan ZHANG Xiaohui LIANG State Key Laboratory of Virtual Reality Technology and Systems Beihang UniversityBeijing 100191China Department of Computer Science and Engineering Shijiazhuang UniversityShijiazhuang 050035China

Background Cumulus clouds are important elements in creating virtual outdoor *** cumulus clouds that have a specific shape is difficult owing to the fluid nature of the ***-based modeling is an efficient method to solve this *** of the complexity of cloud shapes,the task of modeling the cloud from a single image remains in the development *** In this study,a deep learning-based method was developed to address the problem of modeling 3D cumulus clouds from a single *** method employs a three-dimensional autoencoder network that combines the variational autoencoder and the generative adversarial ***,a 3D cloud shape is mapped into a unique hidden space using the proposed ***,the parameters of the decoder are fixed.A shape reconstruction network is proposed for use instead of the encoder part,and it is trained with rendered *** train the presented models,we constructed a 3D cumulus dataset that included 2003D cumulus *** cumulus clouds were rendered under different lighting *** The qualitative experiments showed that the proposed autoencoder method can learn more structural details of 3D cumulus shapes than existing ***,some modeling experiments on rendering images demonstrated the effectiveness of the reconstruction *** The proposed autoencoder network learns the latent space of 3D cumulus cloud *** presented reconstruction architecture models a cloud from a single *** demonstrated the effectiveness of the two models.

关键词： 3D cloud model 3D autoencoder network Generative adversarial network

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：