检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Yang, Dianyi Tang, Jiadong Gao, Yu Yang, Yi Fu, Mengyin School of Automation Beijing Institute of Technology Beijing China State Key Laboratory of Intelligent Control and Decision of Complex System Beijing Institute of Technology Beijing China

Fisheye cameras suffer from image distortion while having a large field of view(LFOV). And this fact leads to poor performance on some fisheye vision tasks. One of the solutions is to optimize the current vision algorithm for fisheye images. However, most of the CNN-based methods and the Transformer-based methods lack the capability of leveraging distortion information efficiently. In this work, we propose a novel patch embedding method called Sector Patch Embedding(SPE), conforming to the distortion pattern of the fisheye image. Furthermore, we put forward a synthetic fisheye dataset based on the ImageNet-1K and explore the performance of several Transformer models on the dataset. The classification top-1 accuracy of ViT and PVT is improved by 0.75% and 2.8% with SPE respectively. The experiments show that the proposed sector patch embedding method can better perceive distortion and extract features on the fisheye images. Our method can be easily adopted to other Transformer-based models. Source code is at https://***/IN2-ViAUn/Sector-Patch-Embedding. Copyright © 2023, The Authors. All rights reserved.

关键词： Embeddings

来源：评论

学校读者我要写书评

暂无评论

Distributed Quadratic Stabilization of Uncertain Linear Multi-Agent systems

Distributed Quadratic Stabilization of Uncertain Linear Mult...

引用

The 24th Chinese control and decision Conference (第24届中国控制与决策学术年会 2012 CCDC)

作者： Zhongkui Li Zhisheng Duan Lihua Xie Xiangdong Liu Key laboratory for Intelligent Control &Decision of Complex Systems School of Automation Beijing I State Key Lab for Turbulence and Complex Systems Department of Mechanics and Aerospace Engineering School of Electrical and ElectronicEngineering Nanyang Technological University 639798 Singapore Key laboratory for Intelligent Control & Decision of Complex Systems School of Automation Beijing

This paper considers the distributed quadratic stabilization problems of uncertain continuous-time linear multiagent systems with undirected communication topologies. It is assumed that the agents have identical nominal dynamics while subject to different norm-bounded parameter uncertainties, leading to weakly heterogeneous multi-agent systems. A distributed controller is proposed, based on the relative states of neighboring agents and a subset of absolute states of the agents. It is shown that the distributed quadratic stabilization problem under such a controller is equivalent to the H∞ control problems of a set of decoupled linear systems having the same dimensions as a single agent. A two-step algorithm is presented to construct the distributed robust controller, which does not involve any conservatism and meanwhile decouples the feedback gain design from the communication topology. Furthermore, the distributed quadratic H∞ control problem of uncertain linear multi-agent systems with external disturbances is discussed, which can be reduced to the scaled H∞ control problems of a set of independent systems whose dimensions are equal to that of a single agent.

关键词： Multi-agent system distributed control robustness H∞ control parameter uncertainty

来源：评论

学校读者我要写书评

暂无评论

CRRS: Concentric Rectangles Regression Strategy for Multi-point Representation on Fisheye Images

arXiv

引用

arXiv 2023年

作者： Wang, Xihan Xu, Xi Gao, Yu Yang, Yi Yue, Yufeng Fu, Mengyin School of Automation Beijing Institute of Technology Beijing China State Key Laboratory of Intelligent Control and Decision of Complex System Beijing Institute of Technology Beijing China

Modern object detectors take advantage of rectangular bounding boxes as a conventional way to represent objects. When it comes to fisheye images, rectangular boxes involve more background noise rather than semantic information. Although multi-point representation has been proposed, both the regression accuracy and convergence still perform inferior to the widely used rectangular boxes. In order to further exploit the advantages of multi-point representation for distorted images, Concentric Rectangles Regression Strategy(CRRS) is proposed in this work. We adopt smoother mean loss to allocate weights and discuss the effect of hyper-parameter to prediction results. Moreover, an accurate pixel-level method is designed to obtain irregular IoU for estimating detector performance. Compared with the previous work for muti-point representation, the experiments show that CRRS can improve the training performance both in accurate and stability. We also prove that multi-task weighting strategy facilitates regression process in this design. Source code is at https://***/IN2-ViAUn/Concentric-Rectangles-Loss. Copyright © 2023, 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

Malaria cell counting diagnosis within large field of view

Malaria cell counting diagnosis within large field of view

引用

International Conference on Digital Image Computing: Techniques and Applications, DICTA 2010

作者： Zou, Li-Hui Chen, Jie Zhang, Juan García, Narciso Education Ministry Key Laboratory of Complex System Intelligent Control and Decision School of Automation Beijing Institute of Technology Beijing China Grupo de Tratamiento de Imágenes E.T.S.I. Telecomunicación Universidad Politécnica de Madrid Madrid Spain

ISBN: (纸本)9780769542713

Malaria is one of the most serious parasitic infections of human. The accurate and timely diagnosis of malaria infection is essential to control and cure the disease. Some image processing algorithms to automate the diagnosis of malaria on thin blood smears are developed, but the percentage of parasitaemia is often not as precise as manual count. One reason resulting in this error is ignoring the cells at the borders of images. In order to solve this problem, a kind of diagnosis scheme within large field of view (FOV) is proposed. It includes three steps. The first step is image mosaicing to obtain large FOV based on space-time manifolds. The second step is the segmentation of erythrocytes where an improved Hough Transform is used. The third step is the detection of nucleated components. At last, it is concluded that the counting accuracy of malaria infection within large FOV is finer than several regular FOVs. © 2010 IEEE.

关键词： Diagnosis

来源：评论

学校读者我要写书评

暂无评论

A Fast Quantity and Position Detection Method Based on Monocular Vision for a Workpieces Counting and Sorting system

A Fast Quantity and Position Detection Method Based on Monoc...

引用

第三十八届中国控制会议

作者： Xin Wang Baokui Li Hongbin Ma Man Luo School of Automation Beijing Institute of Technology State Key Laboratory of Intelligent Control and Decision of Complex Systems Beijing Institute of Technology

With the rapid development of machine vision, many technologies have been applied to the robots for improving the efficiency in the industrial field. This paper concerns the industrial sorting and counting technology problems in a workpieces counting and sorting system, and puts forward a solution using monocular vision. The main process consists of three parts. The rough positioning is accomplished first by using the pixel intensity comparison-based object detection(PICO). Then, image preprocessing and extracting geometric features are established, composing of binarization, morphological operation, optimizing the foreground,finding inner as well as outer contours, and calculating areas. Finally, the center coordinates and categories of workpieces are obtained. We choose nuts and gears as experimental objects, and complete the fast detection. The results of counting nuts and locating gears illustrate that the proposal solution not only has high speed, but also can ensure a high accuracy.

关键词： Monocular Vision the PICO Detector Workpieces Counting Fast Positioning

来源：评论

学校读者我要写书评

暂无评论

A New Modeling Reference Direct Adaptive Sliding Mode control for Electromechanical Actuator

A New Modeling Reference Direct Adaptive Sliding Mode Contro...

引用

Joint 48th IEEE Conference on decision and control and 28th Chinese control Conference

作者： Hao Li Lihua Dou Zhong Su Jie Chen School of Automation and the Key Laboratory of Complex System Intelligent Control and DecisionMinistry of EducationBeijing Institute of Technology School of Automation Beijing Information Science & Technology University

A new model reference direct adaptive sliding mode control(MRDASMC) approach for electromechanical actuator(EMA) is *** with parameter variation,external load disturbance,nonlinear structure and unmodelled dynamics in the system are treated as a lumped *** lumped uncertainty is estimated without the bounds known in advance.A switching function is employed to compensate for the estimated error of the lumped *** the issue of chattering in classical SMC is suppressed as the amplitude of the switching function is tuned with the tracking error by adaption *** adaption law contains a fading factor which can prevent saturation of the control *** stability of the control system is guaranteed by Lyapunov *** results show that the proposed controller can provide favorable performance and is robust to system uncertainties.

关键词： Model reference adaptive sliding mode control Electromechanical actuator Fading factor Saturation Chattering.

来源：评论

学校读者我要写书评

暂无评论

An Adaptive Sliding Mode controller with The Exponential and Power Reaching Law for Discrete systems

An Adaptive Sliding Mode Controller with The Exponential and...

引用

第37届中国控制会议

作者： Xiaofei Zhang Hongbin Ma Zhuang Li Wenchao Zao School of Automation Beijing Institute of Technology State Key Laboratory of Intelligent Control and Decision of Complex Systems Beijing Institute of Technology

With the development of aeronautics and astronautics, the response speed of servo system need be faster. First, In order to improve the dynamic quality of servo system, the exponential and power reaching law, which combines the advantages of the exponential reaching law and the power reaching law, is introduced. Second, the chattering of the sliding mode controller（SMC） with the exponential and power reaching law for discrete systems is investigated. Finally, the adaptive sliding mode controller（ASMC） with the exponential and power reaching law is introduced. The stability of the ASMC with the exponential and power reaching law for discrete systems is analyzed, and the simulation of this approach on one joint of a six degrees of freedom robot is carried out. The experimental results indicate that the ASMC with the exponential and power reaching law is effective in reducing the time of arriving the sliding mode surface. The experimental results also indicate that the ASMC with the exponential and power reaching law may make output error reach zero in a shorter time.

关键词： adaptive sliding mode controller reaching law

来源：评论

学校读者我要写书评

暂无评论

Multimodal Multilevel Attention for Semi-Supervised Skeleton-Based Gesture Recognition

SSRN

引用

SSRN 2024年

作者： Liu, Jinting Gan, Ming-Gang He, Yuxuan Guo, Jia Hu, Kang State Key Laboratory of Intelligent Control and Decision of Complex Systems School of Automation Beijing Institute of Technology Beijing100081 China

Although skeleton-based gesture recognition based on supervised learning has made promising achievements, the reliance on large amounts of annotation for training poses a significant cost. This paper addresses semi-supervised gesture recognition, aiming to learn feature representations effectively from unlabeled data. To resolve this problem, we propose a novel multimodal multilevel attention network to achieve semi-supervised learning. This model uses an attention mechanism to polymerize multimodal and multilevel complementary semantic information of the hand skeleton, designing a multimodal multilevel contrastive loss to measure the similarity of features for semi-supervised learning. Specifically, our method explores the relationships between joint, bone, and motion to learn more discriminative feature representations. Considering the hierarchy of the hand skeleton, the skeleton data is divided into multilevel (i.e., hand-level, finger-level, and joint-level) to capture complementary semantic information. Furthermore, a multimodal contrastive loss measure similarity among multilevel representations. The proposed method outperforms the semi-supervised skeleton-based gesture recognition tasks, according to extensive experimentation on two challenging datasets from SHREC-17 and DHG 14/28 datasets. © 2024, The Authors. All rights reserved.

关键词： Gesture recognition

来源：评论

学校读者我要写书评

暂无评论

Convolutional Neural Network Features Comparison Between Back-Propagation and Extreme Learning Machine

Convolutional Neural Network Features Comparison Between Bac...

引用

第37届中国控制会议

作者： Atmane Khellal Hongbin Ma Qing Fei School of Automation Beijing Institute of Technology State Key Laboratory of Intelligent Control and Decision of Complex Systems Beijing Institute of Technology

Recently deep learning based architectures have been widely deployed in many problems of artificial *** deep learning models, Convolutional Neural Networks（CNN） have been reported in numerous successful applications such as object recognition, and natural language processing. The convolutional neural networks are trained by back-propagating the classification error using the Back-Propagation（BP） algorithm, which requires a large amount of data and slows the training process. To overcome these difficulties, a new fast and accurate approach based on Extreme Learning Machine（ELM） to train any convolutional neural network has been proposed. The developed framework（ELM-CNN） is based on the concept of autoencoding to learn the convolutional filters with biases, by reconstructing the normalized input and the intercept term. In this paper, systematic comparison with traditional back-propagation based training method（BP-CNN） has been made with respect to two aspects qualitative and quantitative. The experimental results on the popular MNIST dataset show that the ELM-CNN algorithm achieves competitive results in terms of generalization performance and up to 16 times faster than the back-propagation based training of CNN.

关键词： Convolutional Neural Network Extreme Learning Machine Features Comparison Visualization

来源：评论

学校读者我要写书评

暂无评论

Adaptive Temporal Aggregation Network for Online Video Visual Relation Detection

SSRN

引用

SSRN 2023年

作者： He, Yuxuan Gan, Ming-Gang Ma, Qianzhao Wang, Chenyu State Key Laboratory of Intelligent Control and Decision of Complex Systems School of Automation Beijing Institute of Technology Beijing100081 China

Nowadays, more and more researchers pay attention to scene perception of artificial robot. Video visual relation detection is an essential task for scene perception but existing methods are all offline methods which are unsuitable for real-time application. Thus, we first introduce online video visual relation detection task which aims at localizing and classifying relations in the current frame. Then, we propose Adaptive Temporal Aggregation Network for online video visual relation detection. In this network, we design content-based adaptive sampling method to sample historical frames which contain dynamic scene information. In order to extract history knowledge from these discrete features, we propose adaptive spatio-temporal activation module which consists of adaptive spatial activation unit and adaptive temporal aggregation unit to extract history knowledge from history memory. After obtaining history knowledge and current state, the proposed attention-based knowledge-state fusion module is used to fuse these dynamic and static visual information. We also construct spatial location features and semantic embedding features to improve final relation representations. Experiments show the effectiveness and efficiency of our method. Moreover, we implement our network on fetch robot and visualize the online video visual relation detecting process. © 2023, The Authors. All rights reserved.

关键词： Chemical activation

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：