检索结果-内蒙古大学图书馆

arXiv 2024年

作者： You, Zeng Wen, Zhiquan Chen, Yaofo Li, Xin Zeng, Runhao Wang, Yaowei Tan, Mingkui The School of Future Technology South China University of Technology Guangzhou China Peng Cheng Laboratory Shenzhen China The School of Software Engineering South China University of Technology Guangzhou China The Artificial Intelligence Research Institute Shenzhen MSU-BIT University China College of Mechatronics and Control Engineering Shenzhen University China The School of Computer Science and Technology Harbin Institute of Technology Shenzhen China

Long video understanding has become a critical task in computer vision, driving advancements across numerous applications from surveillance to content retrieval. Existing video understanding methods suffer from two challenges when dealing with long video understanding: intricate long-context relationship modeling and interference from redundancy. To tackle these challenges, we introduce Fine-Detailed Video Story generation (FDVS), which interprets long videos into detailed textual representations. Specifically, to achieve fine-grained modeling of long-temporal content, we propose a Bottom-up Video Interpretation Mechanism that progressively interprets video content from clips to video. To avoid interference from redundant information in videos, we introduce a Semantic Redundancy Reduction mechanism that removes redundancy at both the visual and textual levels. Our method transforms long videos into hierarchical textual representations that contain multi-granularity information of the video. With these representations, FDVS is applicable to various tasks without any fine-tuning. We evaluate the proposed method across eight datasets spanning three tasks. The performance demonstrates the effectiveness and versatility of our method. Copyright © 2024, The Authors. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Algal Bloom Front Tracking Using an Unmanned Surface Vehicle: Numerical Experiments Based on Baltic Sea Data

Algal Bloom Front Tracking Using an Unmanned Surface Vehicle...

引用

OCEANS 2021: San Diego - Porto

作者： Fonseca, Joana Aguiar, Miguel Sousa, Joao Borges De Johansson, Karl H. KTH Royal Institute of Technology Division of Decision and Control Systems School of Electrical Engineering and Computer Science and Digital Futures StockholmSE-10044 Sweden Portugal

ISBN: (纸本)9780692935590

We consider the problem of tracking moving algal bloom fronts using an unmanned surface vehicle (USV) equipped with a sensor that measures the concentration of chlorophyll a. Chlorophyll a is a green pigment found in plants, and its concentration is an indicator of phytoplankton abundance. Our algal bloom front tracking mission consists of three stages: deployment, data collection, and front tracking. At the deployment stage, a satellite collects an image of the sea from which the location of the front, the reference value for the concentration at this front and, consequently, the appropriate initial position for the USV are determined. At the data collection stage, the USV collects data points to estimate the local algal gradient as it crosses the front. Finally, at the front tracking stage, an adaptive algorithm based on recursive least squares fitting using recent past sensor measures is executed. We evaluate the performance of the algorithm and its sensitivity to measurement noise through MATLAB simulations. We also present an implementation of the algorithm on the DUNE onboard software platform for marine robots and validate it using simulations with satellite model forecasts from Baltic sea data. © 2021 MTS.

关键词： Unmanned surface vehicles

来源：评论

学校读者我要写书评

暂无评论

Door Opening Strategy for Mobile Manipulator with Constrained Configuration 1

引用

7th International Conference on Interactive Collaborative Robotics, ICR 2022

作者： Pushkarev, Daniil Mironov, Konstantin Basharov, Ilya Kichik, Margarita Linok, Sergey Yudin, Dmitry Alhaddad, Muhammad Panov, Aleksandr Moscow Institute of Physics and Technology Institutskiy Per. 9 Moscow Region Dolgoprudny141701 Russia AIRI Artificial Intelligence Research Institute Kutuzovskii Pr 32/1 Moscow121165 Russia Federal Research Center "Computer Sciences and Control" Vavilova 44/2 Moscow119333 Russia

ISBN: (数字)9783031236099

ISBN: (纸本)9783031236082

We address the task of robotic door opening in office environments. This task is important for providing indoor mobility for collaborative mobile manipulators. In our work, we mainly focus on the use of high-level control opportunities and identification of the door parameters from visual and lidar data. We develop a solution, which includes handle recognition, handle twisting, and opening. The position of the handle is identified from stereo images by a neural-network-based method. We divide the opening procedure into two stages: first, handle twisting and slightly opening, and second, wide opening. The first stage is implemented via high-level task-space control of the robotic arm, while the platform is static. The position of the door axis is identified during the slight opening by fitting lidar data to the kinematic model. At the second stage, both the platform and the arm are active. The trajectory of the platform is defined by the model predictive planner in such a way that it avoids pushing the arm into a singular configuration, while the manipulator is operated via high-level impedance control. In our experiments, a mobile manipulator composed from the wheeled platform and the robotic arm was able to open office doors using the proposed approach. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Robotic arms

来源：评论

学校读者我要写书评

暂无评论

Shifting Spotlight for Co-supervision: A Simple yet Efficient Single-branch Network to See Through Camouflage

Shifting Spotlight for Co-supervision: A Simple yet Efficien...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Yang Hu Jinxia Zhang Kaihua Zhang Yin Yuan Jiale Huang Zechao Zhan Xin Wang Key Laboratory of Measurement and Control of CSE School of Automation Southeast University China Advanced Ocean Institute of Southeast University China School of Computer Science Nanjing University of Information Science and Technology China Alibaba Group China

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

Camouflaged object detection (COD) remains a challenging task in computer vision. Existing methods often resort to additional branches for edge supervision, incurring substantial computational costs. To address this, we propose the Co-Supervised Spotlight Shifting Network (CS 3 Net), a compact single-branch framework inspired by how shifting light source exposes camouflage. Our spotlight shifting strategy replaces multi-branch designs by generating supervisory signals that highlight boundary cues. Within CS 3 Net, a Projection Aware Attention (PAA) module is devised to strengthen feature extraction, while the Extended Neighbor Connection Decoder (ENCD) enhances final predictions. Extensive experiments on public datasets demonstrate that CS 3 Net not only achieves superior performance, but also reduces Multiply-Accumulate operations (MACs) by 32.13% compared to state-of-the-art COD methods, striking an optimal balance between efficiency and effectiveness.

关键词： computer vision Image edge detection Object detection Signal processing Feature extraction Acoustics Decoding Computational efficiency Speech processing Light sources

来源：评论

学校读者我要写书评

暂无评论

Decentralized lateral and longitudinal control of vehicle platoons with constant headway spacing

Decentralized lateral and longitudinal control of vehicle pl...

引用

IEEE Conference on Decision and control

作者： Paul Wijnbergen Mark Jeeninga Redmer de Haan Erjen Lefeber School of Electrical Engineering and Computer Science KTH Royal Institute of Technology Stockholm Sweden Department of Automatic Control Lund University Lund Sweden Department of Mechanical Engineering Eindhoven University of Technology Eindhoven The Netherlands

The formation of platoons, where groups of vehicles follow each other at close distances, has the potential to increase road capacity. In this paper, a decentralized control approach is presented that extends the well-known constant headway vehicle following approach to the two-dimensional case, i.e., lateral control is included in addition to the longitudinal control. The presented control scheme employs a direct vehicle following approach where each vehicle in the platoon is responsible for following the directly preceding vehicle according to a nonlinear spacing policy. The proposed constant headway spacing policy is motivated by an approximation of a delay-based spacing policy and results in a generalization of the constant headway spacing policy to the two-dimensional case. By input-output linearization, necessary and sufficient conditions for the tracking of the nonlinear spacing policy are obtained, which motivate the synthesis of the lateral and longitudinal controllers of each vehicle in the platoon. By deriving an internal state representation of the follower vehicle and showing input-to-state stability, the internal dynamics for each leader-follower subsystem are shown to be well-behaved in case the leader drives in steady state conditions (i.e., the leader vehicle's trajectory is unexcited). The results are illustrated by a simulation.

关键词：

来源：评论

学校读者我要写书评

暂无评论

IR Net: information restriction and information recovery for accurate binary neural networks

引用

Neural Computing and Applications 2023年第19期35卷 14449-14464页

作者： Xue, Ping Lu, Yang Chang, Jingfei Wei, Xing Wei, Zhen School of Computer Science and Information Engineering Hefei University of Technology Hefei China Anhui Mine IOT and Security Monitoring Technology Key Laboratory Hefei China Engineering Research Center of Safety Critical Industrial Measurement and Control Technology Ministry of Education Hefei University of Technology Hefei China Intelligent Manufacturing Institute of Hefei University of Technology Hefei China

Weight and activation binarization can efficiently compress deep neural networks and accelerate model inference, but they cause severe accuracy degradation. Existing optimization methods for binary neural networks (BNNs) focus on fitting full-precision networks to reduce quantization errors and suffer from a tradeoff between accuracy and efficiency. In contrast, considering information loss and the mismatch between model capacity and input information quantity caused by network binarization, we propose Information Restriction and Information Recovery Network (IR $$^2$$ Net) to stimulate the potential of BNNs and achieve improved network accuracy by restricting the input information and recovering feature information. The proposed approach includes (1) information restriction, which evaluates the feature information extracted from the input by a BNN, discards some of the information it cannot focus on, and reduces the amount of the input information to match the model capacity; and (2) information recovery: due to the information loss incurred during forward propagation, the extracted feature information of the network is not sufficient for supporting accurate classification. Shallow feature maps with richer information are selected, and these feature maps are fused with the final feature maps to recover the extracted feature information and further enhance the model capacity to match the amount of input information. In addition, the computational cost is reduced by streamlining the information recovery method to strike a better tradeoff between accuracy and efficiency. Experimental results demonstrate that our approach still achieves comparable accuracy even with a $$\sim$$ 10x floating-point operations (FLOPs) reduction for ResNet-18. The models and code are available at https://***/pingxue-hfut/IR2Net .

关键词：

来源：评论

学校读者我要写书评

暂无评论

Selection of Cancer Biomarkers from Microarray Gene Expression Data Utilizing the Bi-Objective Optimization Method

Selection of Cancer Biomarkers from Microarray Gene Expressi...

引用

2023 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2023

作者： Jiang, Hanjing Lin, Jianqing Zhu, Huachao Huang, Yabing Huazhong University of Science and Technology Key Lab. of Image Information Processing and Intelligent Control of Education Ministry of China Institute of Artificial Intelligence School of Artificial Intelligence and Automation Wuhan430074 China Wuhan University School of Computer Science Wuhan430064 China Renmin Hospital of Wuhan University Department of Pathology Wuhan430060 China

ISBN: (纸本)9798350337488

Cancer-associated biomarker genes play an indispensable role in the intricate tapestry of cancer development and manifestation. The expression of biomarkers in different types of tumor cells has beneficial implications for shedding light on the development of various cancers, guiding clinical diagnosis, and treatment. Microarray technology enables the expression levels of thousands of genes in samples to be sequenced simultaneously. However, sparse and high-dimensional microarray data present a formidable challenge in identifying biomarker genes. This study presents EREF-NSGA2, a novel method for cancer biomarker selection from microarray data, employing a hybrid gene selection approach. Firstly, the combination of the wrapper and embedded gene selection methods is proposed to filter the microarray data, which efficiently decreases the search space of the algorithm. After that, the improved NSGA-II algorithm is used to search the genes subset obtained from the previous step to reach the optimal subset of cancer biomarker genes. The proposed EREF-NSGA2 is compared with other reported methods on six cancer benchmark gene expression datasets. A detailed biological analysis is performed to analyze the relationship between the selected genes and the cancer data sets they belong to. To summarize, EREF-NSGA2 proves its effectiveness in selecting a feature subset comprising the fewest genes while maintaining the highest classification accuracy. © 2023 IEEE.

关键词： Bi-objective optimization Cancer biomarker Gene selection NSGA-II

来源：评论

学校读者我要写书评

暂无评论

Edge Accelerated Robot Navigation With Collaborative Motion Planning

arXiv

引用

arXiv 2023年

作者： Li, Guoliang Han, Ruihua Wang, Shuai Gao, Fei Eldar, Yonina C. Xu, Chengzhong Department of Computer and Information Science University of Macau 999078 China The Department of Computer Science The University of Hong Kong Hong Kong The Shenzhen Institute of Advanced Technology Chinese Academy of Sciences Shenzhen518055 China The State Key Laboratory of Industrial Control Technology The Huzhou Institute Zhejiang University Zhejiang 310027 China The Faculty of Mathematics and Computer Science Weizmann Institute of Science Rehovot7610001 Israel

Low-cost distributed robots suffer from limited onboard computing power, resulting in excessive computation time when navigating in cluttered environments. This paper presents Edge Accelerated Robot Navigation (EARN), to achieve real-time collision avoidance by adopting collaborative motion planning (CMP). As such, each robot can dynamically switch between a conservative motion planner executed locally to guarantee safety (e.g., path-following) and an aggressive motion planner executed non-locally to guarantee efficiency (e.g., overtaking). In contrast to existing motion planning approaches that ignore the interdependency between low-level motion planning and high-level resource allocation, EARN adopts model predictive switching (MPS) that maximizes the expected switching gain with respect to robot states and actions under computation and communication resource constraints. The MPS problem is solved by a tightly-coupled decision making and motion planning framework based on bilevel mixed-integer nonlinear programming and penalty dual decomposition. We validate the performance of EARN in indoor simulation, outdoor simulation, and real-world environments. Experiments show that EARN achieves significantly smaller navigation time and higher success rates than state-of-the-art navigation approaches. Copyright © 2023, The Authors. All rights reserved.

关键词： Motion planning

来源：评论

学校读者我要写书评

暂无评论

Integrating Static Code Analysis Toolchains

arXiv

引用

arXiv 2024年

作者： Kern, Matthias Erata, Ferhat Iser, Markus Sinz, Carsten Loiret, Frederic Otten, Stefan Sax, Eric FZI Research Center for Information Technology Karlsruhe Germany Karlsruhe Institute of Technology Institut für Theoretische Informatik Karlsruhe Germany KTH Royal Institute of Technology Embedded Control Systems Stockholm Sweden Yale University Department of Computer Science New Haven United States UNIT Information Technologies Research & Development Izmir Turkey

This paper proposes an approach for a tool-agnostic and heterogeneous static code analysis toolchain in combination with an exchange format. This approach enhances both traceability and comparability of analysis results. State of the art toolchains support features for either test execution and build automation or traceability between tests, requirements and design information. Our approach combines all those features and extends traceability to the source code level, incorporating static code analysis. As part of our approach we introduce the "ASSUME Static Code Analysis tool exchange format" that facilitates the comparability of different static code analysis results. We demonstrate how this approach enhances the usability and efficiency of static code analysis in a development process. On the one hand, our approach enables the exchange of results and evaluations between static code analysis tools. On the other hand, it enables a complete traceability between requirements, designs, implementation, and the results of static code analysis. Within our approach we also propose an OSLC specification for static code analysis tools and an OSLC communication framework. Copyright © 2024, The Authors. All rights reserved.

关键词： Static analysis

来源：评论

学校读者我要写书评

暂无评论

Learning Reduced-Order Linear Parameter-Varying Models of Nonlinear Systems

arXiv

引用

arXiv 2023年

作者： Koelewijn, Patrick J.W. Singh, Rajiv Seiler, Peter Tóth, Roland Sioux Technologies B.V. Eindhoven Netherlands Control Systems Group Eindhoven University of Technology Eindhoven Netherlands The MathWorks Inc. Natick United States Department of Electrical Engineering and Computer Science University of Michigan Ann Arbor United States Systems and Control Laboratory HUN-REN Institute for Computer Science and Control Budapest Hungary

In this paper, we consider the learning of a Reduced-Order Linear Parameter-Varying Model (ROLPVM) of a nonlinear dynamical system based on data. This is achieved by a two-step procedure. In the first step, we learn a projection to a lower dimensional state-space. In step two, an LPV model is learned on the reduced-order state-space using a novel, efficient parameterization in terms of neural networks. The improved modeling accuracy of the method compared to an existing method is demonstrated by simulation examples. Copyright © 2023, The Authors. All rights reserved.

关键词： Nonlinear dynamical systems

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：