检索结果-内蒙古大学图书馆

SiamS3C: spatial-channel cross-correlation for visual tracking with centerness-guided regression

Multimedia Systems 2024年第5期30卷 1-14页

作者： Zhang, Jianming Chen, Wentao He, Yufan Kuang, Li-Dan Sangaiah, Arun Kumar School of Computer and Communication Engineering Changsha University of Science and Technology Changsha410076 China International Graduate School of AI National Yunlin University of Science and Technology Douliu64002 Taiwan

Visual object tracking can be divided into the object classification and bounding-box regression tasks, but only one sharing correlation map leads to inaccuracy. Siamese trackers compute correlation map by cross-correlation operation with high computational cost, and this operation performed either on channels or in spatial domain results in weak perception of the global information. In addition, some Siamese trackers with a centerness branch ignore the associations between the centerness branch and the bounding-box regression branch. To alleviate these problems, we propose a visual object tracker based on Spatial-Channel Cross-Correlation and Centerness-Guided Regression. Firstly, we propose a spatial-channel cross-correlation module (SC3M) that combines the search region feature and the template feature both on channels and in spatial domain, which suppresses the interference of distractors. As a lightweight module, SC3M can compute dual independent correlation maps inputted to different subnetworks. Secondly, we propose a centerness-guided regression subnetwork consisting of the centerness branch and the bounding-box regression branch. The centerness guides the whole regression subnetwork to enhance the association of two branches and further suppress the low-quality predicted bounding boxes. Thirdly, we have conducted extensive experiments on five challenging benchmarks, including GOT-10k, VOT2018, TrackingNet, OTB100 and UAV123. The results show the excellent performance of our tracker and our tracker achieves real-time requirement at 48.52 fps. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2024.

关键词： Object tracking

来源：评论

学校读者我要写书评

暂无评论

Depth Estimation with Tilted Optics Using Split Color Filter Aperture 4

Depth Estimation with Tilted Optics Using Split Color Filter...

引用

4th IEEE International Conference on Electrical, computer, and Energy Technologies, ICECET 2024

作者： Fukino, Aoi Ikeoka, Hiroshi Hamamoto, Takavuki Graduate School of Engineering Fukuyama University Information Processing Engineering Course Hiroshima Japan Fukuyama University Faculty of Engineering Department of Computer Science Hiroshima Japan

ISBN: (纸本)9798350395914

Camera-based depth estimation methods are useful for various automobile applications. For example, depth estimation methods based on stereo cameras triangulate the depth between objects and cameras. However, they are susceptible to baseline distortion owing to vibrations, aging, and thermal expansion. On the contrary, depth estimation methods based on a monocular camera, such as depth-from-defocus, use a focusing mechanism or spectroscopic optics to calculate the depth from the difference of defocus images. However, because the change in sharpness, which is the inverse of defocus, is slightly in the depth direction, it is unsuitable for deep-range estimation. Therefore, we are developing depth-estimation methods that use a monocular camera with special focus images obtained through tilted optics, which have no baseline distortion problem and have a deep estimation range. When tilted optics are used, the plane of sharp focus (POF) appears tilted, while the depth of field (DOF) lies and increases toward the depth direction. Moreover, to allow for the use of only one image-sensor, a one-shot, we studied a depth estimation method using tilted optics with two different aperture sizes for each color wavelength by introducing concentric color filter apertures. However, as with conventional methods, the Gaussian sharpness model does not strictly fit to the actual sharpness, resulting in reduced estimation accuracy. In this study, we propose a method for high-speed and high-accuracy depth estimation that involves adding a split color filter aperture to a tilted optic and using a different sharpness model for each DOF side. © 2024 IEEE.

关键词： Stereo image processing

来源：评论

学校读者我要写书评

暂无评论

Near-Far Field Beamforming for Holographic Multiple-Input Multiple-Output

引用

Journal of communications and Information Networks 2023年第2期8卷 99-110页

作者： Yutong Zhang Boya Di Hongliang Zhang Lingyang Song State Key Laboratory of Advanced Optical Communication Systems and Networks School of ElectronicsPeking UniversityBeijing 100871China School of Electronic and Computer Engineering Peking University Shenzhen Graduate SchoolShenzhen 518055China

Recently,holographic multiple-input multiple-output(HMIMO)has motivated its potential use to support high-capacity data transmission with spatially quasi-continuous *** a practical instance of HMIMO,reconfigurable refractive surfaces(RRSs)equipped with numerous metamaterial elements are utilized as antennas by refracting incident signals from signal *** this paper,we investigate a multi-user communication system with an RRS deployed as the base station(BS)’s transmit *** mitigate the high overhead of accurate channel state information(CSI)acquisition,the codebook design and beam training are employed to perform *** the large scale of RRS,users are likely to be randomly distributed in both the near and far fields around the BS,which is unknown in *** considering radiation characteristics in both fields,a near-far field codebook is designed to be applicable to all users,regardless of their *** reduce overhead,a multi-user beam training is proposed to serve all users simultaneously by enhancing each codeword capable of covering multiple *** a general case that includes users in both fields,simulation results indicate that,without prior knowledge of user distribution,the proposed scheme outperforms state-of-the-art ones in terms of sum rate and overhead.

关键词： near-far field codebook design RRSs multi-user beam training

来源：评论

学校读者我要写书评

暂无评论

Analysis of the performance of robotic assembly systems considering the usage of tools or jigs for the task-board task in the WRS 2018

引用

Advanced Robotics 2023年第20期37卷 1329-1346页

作者： Dobashi, Hiroki Ogawa, Koki Shibata, Mizuho Uemura, Wataru Yokokohji, Yasuyoshi Faculty of Systems Engineering Wakayama University Wakayama Japan Graduate School of Systems Engineering Wakayama University Wakayama Japan Department of Robotics Faculty of Engineering Kindai University Higashi-Hiroshima Japan Electronics Information and Communication Engineering Course Faculty of Advanced Science and Technology Ryukoku University Otsu Japan Department of Mechanical Engineering Graduate School of Engineering Kobe University Kobe Japan

A task-board task is a task in which robots perform a set of fundamental assembly operations related to the assembly of a specific product such as inserting a part into another part, tightening bolts and nuts, attaching flexible parts, etc. In the task-board task competition in the industrial robotics category of the World Robot Summit (WRS) 2018, 15 teams from around the world performed the task. In our previous work, we observed recorded videos of the competition frame by frame and manually investigated the performance of robotic systems of the top four teams, focusing only on whether robots contact target parts. In this paper, we introduce frame sets to represent mainly contact states between robots, tools, or jigs and target parts to be handled for such a frame-by-frame analysis of the recorded videos, and based on the frame sets, we analyze the performance of robotic assembly systems of other teams as well as the top four teams in the task-board task competition in the WRS 2018, considering the usage of tools or jigs. © 2023 Informa UK Limited, trading as Taylor & Francis Group and The Robotics Society of Japan.

关键词： Robotic assembly

来源：评论

学校读者我要写书评

暂无评论

Maneuverability Improvement by Feedback Compensation of a Physically Human-Interacting Robot with Variable Damping Control

Maneuverability Improvement by Feedback Compensation of a Ph...

引用

18th IEEE International Conference on Advanced Motion Control, AMC 2024

作者： Songthumjitti, Narawich Inara, Takpeshi Graduate School of Tokai University Course of Electrical and Electronic Engineering Kanagawa Japan Tokai University Department of Applied Computer Engineering Kanagawa Japan

ISBN: (纸本)9798350342819

Manufacturing workflows can be enhanced using the implementation of robots;however, it is limited to repetitive tasks, and a human operator is still needed to gain flexibility in the robot's movement. However, human-robot interaction can cause instability due to the robot's characteristics. The typical solution is to decrease the admittance parameter, but it also reduces the assisting effectiveness. Variable damping is another way to counter this problem, but if the instability problem isn't solved yet, the effect can be difficult to perceive. Our study aims to implement feedback compensation in the robot system to reduce the effect of actuator dead time and insufficient structural stiffness. The results prove that the implementation of feedback compensation can greatly improve the maneuverability of the robot by decreasing the force required during the traveling phase and increasing accuracy in the positioning phase, resulting in a task that can be completed with less effort and a shorter operation time. © 2024 IEEE.

关键词： Human robot interaction

来源：评论

学校读者我要写书评

暂无评论

Improved encoding structure and decoding algorithms for spinal codes

引用

Journal of Systems engineering and Electronics 2022年第4期33卷 856-866页

作者： HUANG Wensha WANG Lina School of Computer and Communication Engineering University of Science and Technology BeijingBeijing 100083China Shunde Graduate School University of Science and Technology BeijingBeijing 100083China

To improve the error correction performance, an innovative encoding structure with tail-biting for spinal codes is designed. Furthermore, an adaptive forward stack decoding(A-FSD) algorithm with lower complexity for spinal codes is proposed. In the A-FSD algorithm, a flexible threshold parameter is set by a variable channel state to narrow the scale of nodes accessed. On this basis, a new decoding method of AFSD with early termination(AFSD-ET) is further proposed. The AFSD-ET decoder not only has the ability of dynamically modifying the number of stored nodes, but also adopts the early termination criterion to curtail complexity. The complexity and related parameters are verified through a series of simulations. The simulation results show that the proposed spinal codes with tail-biting and the AFSD-ET decoding algorithms can reduce the complexity and improve the decoding rate without sacrificing correct decoding performance.

关键词： spinal code tail-biting encoding lower decoding complexity early termination strategy

来源：评论

学校读者我要写书评

暂无评论

Effect of Graphite Addition on Microstructure, Mechanical Properties and Thermal Properties of Injection Molded AZ91D Alloy+1

引用

Materials Transactions 2024年第4期65卷 374-380页

作者： Hideshima, Yasutoshi Maeda, Fumiya Fukuta, Tadao Ozaki, Koichi Course of Advanced Systems Engineering Graduate School of Computer Science and Systems Engineering Okayama Prefectural University Soja719-1197 Japan Process Technology Development Department Technology Development Division Seiko Epson Corp. Suwa399-0211 Japan Department of Systems Engineering Faculty of Computer Science and Systems Engineering Okayama Prefectural University Soja719-1197 Japan

Magnesium chips were coated with a high concentration of graphite using a binder and were used as the raw material for injection molding. The microstructure of the magnesium injection-molded product with added graphite exhibited a dispersion of needle-like graphite particles. No significant voids were observed at the interfaces between the graphite and the matrix. The addition of more than 0.5 mass% graphite decreased the proof stress and tensile strength of the injection-molded products. The Young’s modulus of the graphite-added products tended to decrease with an increase in the graphite content, which is consistent with the lower limit of the rule of mixtures. The thermal conductivity of the 6.9 mass% graphite-added product increased compared with that of the AZ91D magnesium alloy and the coefficient of linear thermal expansion decreased. Both values are within a range that satisfied the rules of mixtures. © 2024 The Japan Institute of Metals and Materials.

关键词： Thermal expansion

来源：评论

学校读者我要写书评

暂无评论

Distance Characteristics of Power Absorption Ratio of the Skin Based on Sommerfeld's Theory 7

Distance Characteristics of Power Absorption Ratio of the Sk...

引用

7th IEEE Global Electromagnetic Compatibility Conference, GEMCCON 2023

作者： Sato, Ken Kamimura, Yoshitsugu National Institute of Technology Hachinohe College Electrical and Computer Engineering Course Hachinohe Japan Graduate School of Engineering Utsunomiya University Division of Engineering and Agriculture Utsunomiya Japan

ISBN: (纸本)9798350396935

The dependence of the amount of power absorbed by the skin on the distance from the small antenna was analyzed based on the theory of Sommerfeld and Renner. The results showed good agreement between the theoretical and numerical values, with all power absorbed at small separation distance and converging to about 20 % at larger distance. © 2023 IEEE.

关键词： Antennas

来源：评论

学校读者我要写书评

暂无评论

Automatic Identification of Daily Life Activities and Prediction of Falling Accidents Caused by Behavior 25th

Automatic Identification of Daily Life Activities and Pred...

引用

25th International Conference on Human-computer Interaction, HCII 2023

作者： Hayashi, Seiji Nakajima, Akira Department of Electronics and Computer Systems Faculty of Engineering Takushoku University Tokyo Japan Mechanical and Electronic Systems Course Graduate School of Engineering Takushoku University Tokyo Japan

ISBN: (纸本)9783031359910

In the present research, we developed an application that measures acceleration by a smartphone located at the waist, along with a deep learning system that automatically detects six behavioral activities (walking, running, standing, walking downstairs, walking upstairs, and falling). Specifically, we use a human activity database provided by the University of California–Irvine (UCI) to create a deep learning model with the UCI data. Next, transfer learning was applied to the learning model for the measurement data obtained by the acceleration acquisition application developed in this research. Using Sony’s Neural Network Console development environment as a deep learning tool, in recognition experiments, our proposed method for automatic identification of activities and prediction of falling accidents caused by behavior achieved a high accuracy rate of almost 95%. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

The Effectiveness of CNN-Based Models in Macroscopic Motion Direction Detection 6

The Effectiveness of CNN-Based Models in Macroscopic Motion ...

引用

6th International Conference on Robotics and computer Vision, ICRCV 2024

作者： Qiu, Zhiyu Chen, Tianqi Hua, Yuxiao Todo, Yuki Graduate School of Natural Science & Technology Kanazawa University Division of Electrical Engineering and Computer Science Kanazawa Japan Kanazawa University Faculty of Electrical Information and Communication Engineering Kanazawa Japan

ISBN: (纸本)9798331527426

Convolutional Neural Networks (CNNs) have revolutionized the field of computer vision, offering significant advancements in image and video recognition, classification, and segmentation tasks. Leveraging hierarchical pattern recognition capabilities, CNNs have demonstrated superior performance over traditional machine learning methods. One particularly critical application of CNNs is in motion direction detection, which is essential for enhancing security in surveillance systems, ensuring safe navigation in autonomous vehicles, and providing valuable insights in sports analytics. In this study, our primary goal was to investigate how different types of input data impact the accuracy of motion direction detection. We applied five distinct preprocessing methods to test the effect of input data on accuracy. Furthermore, we conducted experiments in various object and background scenarios to evaluate adaptability of simple 4-layer CNN, EfficientNetBO and ResNet50 in detecting motion direction. Finally, we tested the noise resistance of these three models. The results indicate that CNNs are capable of extracting the macroscopic motion direction of objects in images. Our study highlights the potential of CNN-based models for efficient and effective motion direction detection. © 2024 IEEE.

关键词： 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：