检索结果-内蒙古大学图书馆

13th IEEE International Conference on Multisensor Fusion and Integration for intelligent Systems, MFI 2017

ISBN: (纸本)9781509060641

The proceedings contain 108 papers. The topics discussed include: intelligent grasping of various moving objects based upon visual servoing;a patient-to-CT registration method based on spherical unwrapping and H-K curvature descriptors for surgical navigation system;online reliability assessment and reliability-aware fusion for ego-lane detection using influence diagram and Bayes filter;multirotor UAV state prediction through multi-microphone side-channel fusion;global point-to-hyperplane ICP: local and global pose estimation by fusing color and depth;on weak points of the ellipsoidal intersection fusion;facility activity inference using networks of radiation detectors based on SPRT;entropy-based abnormal activity detection fusing RGB-D and domotic sensors;discretization of SO(3) using recursive tesseract subdivision;a deep neural network approach to fusing vision and heteroscedastic motion estimates for low-SWaP robotic applications;relevance and redundancy as selection techniques for human-autonomy sensor fusion;underwater terrain navigation using standard sea charts and magnetic field maps;on state estimation and fusion with elliptical constraints;a general framework for data fusion and outlier removal in distributed sensor networks;and wide bandwidth nonlinear 2-DOF energy harvester: modeling and parameters selection.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Estimation of local and global superpixel covariance for salient object detection in low contrast images 2017

Estimation of local and global superpixel covariance for sal...

引用

9th International Conference on Machine Learning and Computing, ICMLC 2017

作者： Mu, Nan Xu, Xin Zhang, Xiaolong School of Computer Science and Technology Wuhan University of Science and Technology Wuhan Hubei430065 China Hubei Province Key Laboratory of Intelligent Information Processing and Real-Time Industrial Systems Wuhan University of Science and Technology Wuhan Hubei430065 China

ISBN: (纸本)9781450348171

Salient object detection has become a hot topic in computer vision as it can substantially facilitate a wide range of applications. Conventional salient object detection models primarily rely on low-level image features, which may face great difficulties in low lighting scenarios. This paper proposes to estimate the saliency of low contrast images via covariance features. The input image is firstly decomposed into superpixel regions to estimate their covariances. Then, the local and global image saliency can be calculated using the covariance features respectively. Finally, a graph-based diffusion process is performed to refine the saliency maps. Extensive experiments have been conducted to evaluate the performance of the proposed model against eleven state-of-the-art models on five benchmark datasets and a nighttime image dataset. © 2017 ACM.

关键词： Superpixels

来源：评论

学校读者我要写书评

暂无评论

Combining Depth and Intensity Images to Produce Enhanced Object Detection for Use in a Robotic Colony 1

引用

18th Annual Conference on Towards Autonomous robotics (TAROS)

作者： Balding, Steven Davis, Darryl N. Univ Hull Fac Engn & Comp Sci Kingston Upon Hull N Humberside England

ISBN: (数字)9783319641072

ISBN: (纸本)9783319641072;9783319641065

Robotic colonies that can communicate with each other and interact with their ambient environments can be utilized for a wide range of research and industrial applications. However amongst the problems that these colonies face is that of the isolating objects within an environment. Robotic colonies that can isolate objects within the environment can not only map that environment in detail, but interact with that ambient space. Many object recognition techniques exist, however these are often complex and computationally expensive, leading to overly complex implementations. In this paper a simple model is proposed to isolate objects, these can then be recognize and tagged. The model will be using 2D and 3D perspectives of the perceptual data to produce a probability map of the outline of an object, therefore addressing the defects that exist with 2D and 3D image techniques. Some of the defects that will be addressed are;low level illumination and objects at similar depths. These issues may not be completely solved, however, the model provided will provide results confident enough for use in a robotic colony.

关键词： Anchoring Robotic vision Sobel Depth map Robotic colony

来源：评论

学校读者我要写书评

暂无评论

FIST-Toy: A benchmark dataset for toy component classification

FIST-Toy: A benchmark dataset for toy component classificati...

引用

2019 IEEE International Conference on Real-Time Computing and robotics, RCAR 2019

作者： Qi, Chongchong Gong, Libo Cheng, Jun Tao, Dapeng FIST LAB School of Information Science and Engineering Yunnan University Kunming Yunnan650091 China Guangdong Provincial Key Laboratory of Robotics and Intelligent System Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences Shenzhen518055 China Yunnan Rural Science and Technology Service Center Kunming Yunnan650000 China Chinese University of Hong Kong Hong Kong999077 Hong Kong

ISBN: (纸本)9781728137261

As a comprehensive technology of machinery, electronics and computer vision, industrial automation technology has been attracting extensive attention and developed rapidly recently years. The refinement of target classification plays an important role in the industrial automation, therefore, many computer vision based target classification methods have been proposed. However, there is few real industrial dataset can be used to evaluating the target classification algorithms. In this paper, we construct a novel toys dataset, named FIST-Toy, including 7411 images of 31 categories. Base on the proposed dataset, we conduct a comprehensive study of the state-of-the-art pretraining target classification models qualitatively and quantitatively. In addition, we propose a weight voting based classification neural network. By comparing with AlexNET, VggNet, Inception and ResNet, the proposed model outperforms in fine-grading. We evaluate the FIST-Toy dataset using various deep learning models. And the experimental results show that the FIST-Toy dataset provides a benchmark to evaluate the performance of different methods for target classification. © 2019 IEEE.

关键词： computer vision

来源：评论

学校读者我要写书评

暂无评论

Remaining Useful Battery Life Prediction for UAVs based on Machine Learning

引用

IFAC-PapersOnLine 2017年第1期50卷 4727-4732页

作者： Mansouri S.S. Karvelis P. Georgoulas G. Nikolakopoulos G. Robotics Group Control Engineering Division of the Department of Computer Electrical and Space Engineering Luleå University of Technology Luleå SE-97187 Sweden Laboratory of Knowledge and Intelligent Computing Department of Computer Engineering Technological Educational Institute of Epirus Arta Greece

Unmanned Aerial Vehicles are becoming part of many industrial applications. The advancements in battery technologies played a crucial part for this trend. However, no matter what the advancements are, all batteries have a fixed capacity and after some time drain out. In order to extend the flying time window, the prediction of the time that the battery will no longer be able to support a flying condition is crucial. This in fact can be cast as a standard Remaining Useful Life prognostic problem, similarly encountered in many fields. In this article, the problem of Remaining Useful Life estimation of a battery, under different flight conditions, is tackled using four machine learning techniques: a linear sparse model, a variant of support vector regression, a multilayer perceptron and an advanced tree based algorithm. The efficiency of the overall proposed machine learning techniques, in the field of batteries prognostics, is evaluated based on multiple experimental data from different flight conditions. © 2017

关键词： Battery Machine Learning Prediction Remaining Useful Life UAVs

来源：评论

学校读者我要写书评

暂无评论

RGB-D-Based Object Recognition using Multimodal Convolutional Neural Networks: A Survey

引用

IEEE ACCESS 2019年 7卷 43110-43136页

作者： Gao, Mingliang Jiang, Jun Zou, Guofeng John, Vijay Liu, Zheng Shandong Univ Technol Sch Elect & Elect Engn Zibo 255000 Peoples R China Univ British Columbia Fac Appl Sci Vancouver BC V1V 1V7 Canada Southwest Petr Univ Sch Comp Sci & Technol Chengdu 610500 Sichuan Peoples R China Sichuan Univ Sci & Engn Sch Comp Sci Zigong 643000 Peoples R China Toyota Technol Inst Intelligent Informat Proc Lab Nagoya Aichi 4688511 Japan

Object recognition in real-world environments is one of the fundamental and key tasks in computer vision and robotics communities. With the advanced sensing technologies and low-cost depth sensors, the high-quality RGB and depth images can be recorded synchronously, and the object recognition performance can be improved by jointly exploiting them. RGB-D-based object recognition has evolved from early methods that using hand-crafted representations to the current state-of-the-art deep learning-based methods. With the undeniable success of deep learning, especially convolutional neural networks (CNNs) in the visual domain, the natural progression of deep learning research points to problems involving larger and more complex multimodal data. In this paper, we provide a comprehensive survey of recent multimodal CNNs (MMCNNs)-based approaches that have demonstrated significant improvements over previous methods. We highlight two key issues, namely, training data deficiency and multimodal fusion. In addition, we summarize and discuss the publicly available RGB-D object recognition datasets and present a comparative performance evaluation of the proposed methods on these benchmark datasets. Finally, we identify promising avenues of research in this rapidly evolving field. This survey will not only enable researchers to get a good overview of the state-of-the-art methods for RGB-D-based object recognition but also provide a reference for other multimodal machine learning applications, e.g., multimodal medical image fusion, audio-visual speech recognition, and multimedia retrieval and generation.

关键词： Convolutional neural network multimodal fusion object recognition RGB-D survey

来源：评论

学校读者我要写书评

暂无评论

Hardware-in-the-loop evaluation of an LPV sliding mode fixed control allocation scheme on the MuPAL-α research aircraft 1

Hardware-in-the-loop evaluation of an LPV sliding mode fixed...

引用

1st Annual IEEE Conference on Control Technology and applications

作者： Chen, L. Alwi, H. Edwards, C. Sato, M. Univ Exeter Coll Engn Math & Phys Sci Exeter EX4 4QF Devon England Japan Aerosp Explorat Agcy Mitaka Tokyo 1810015 Japan

ISBN: (纸本)9781509021826

This paper develops a sliding mode fault tolerant control scheme based on an LPV system representation of the plant. The scheme involves a control allocation component, which is capable of fully utilizing the available actuators in the face of actuator faults. In this paper, information about the actuator faults is assumed not to be available and therefore a fixed control allocation structure is utilised in the event of faults. The proposed scheme is validated using the Japanese Aerospace Exploration Agency's Multi-Purpose Aviation Laboratory (MuPAL-alpha) research aircraft. This paper describes initial hardware-in-the-loop (HIL) tests which serve as a precursor to upcoming real flight tests. The validation results show good lateral-directional state tracking performance in the fault free case with no visible performance degradation in the presence of (aileron) faults. Successful HIL tests demonstrate the potential of the proposed scheme which will be flight tested later this year.

关键词： an LPV MuPAL-α plant degradation of performance RESEARCH AIRCRAFT HARDWARE IN THE LOOP plants faults Sliding mode Scheme computer hardware high intensity lighting actuators

来源：评论

学校读者我要写书评

暂无评论

Semantic Image Recognition Methods by using Deep Learning

Semantic Image Recognition Methods by Using Deep Learning

引用

作者： Dang Hoang Anh Waseda University

学位级别：博士

Artificial intelligence (AI) used to be widely perceived as the field that studies intelligent agent (IA) [1–4]. After the deep learning (DL) breakthrough of AlexNet in 2012 [5], the term AI was frequently used by media, marketers, and academia alike in a much broader sense. In the annual report, the Stanford University's Institute for HumanCentered Artificial Intelligence (HAI) includes computer vision (CV), Pattern Recognition, Computational Linguistics (CL), robotics, as the subcategories of AI[6]. In point of fact, AI is generally defined as intelligence demonstrated by machines. As such, machine learning and its deep learning subcategory are also parts of IA in computer science. Since the publication of AlexNet in 2012, DL has been adopted by many branches of science. However, DL sees the most success in the fields of natural language processing (NLP), CV, and IA. In the field of CV, humanlevel performance in the task of image classification has been achieved by multiple convolutional neural network (CNN) models in 2016. As a result, the ImageNet [7] challenge discontinued in 2017. The end of the ImageNet challenge marks the new chapter in the field. The focus of the research community was shifted to more challenging topics. Some of the notable topics include generative network, semantic segmentation, activity recognition, and visual question answering (VQA). As a whole, the targets of DL researches have become highly semantic. Significant progress is made among all of the fields in recent years [6]. However, the pen etration of AI in everyday life is still limited. Like any other technology, the adoption of DL based AI in the general consumer market lags behind the adoption speed of the industrial mar ket. Not to mention that the adaption speed of the industrial market itself is also lagging behind the research a considerable amount of time. To be adopted by the industrial and general consumer market, further research must be done to adapt and improve the

关键词：

来源：评论

学校读者我要写书评

暂无评论

Recovering the Manhattan Frame from A Single RGB-D Image by using Orientation Relevance 29

Recovering the Manhattan Frame from A Single RGB-D Image by ...

引用

第29届中国控制与决策会议

作者： Zhiqiu Wu Liang Wang College of Automation Faculty of Information TechnologyBeijing University of Technology Beijing Key Laboratory of Computational Intelligence and Intelligent System

ISBN: (纸本)9781509046584

Most man-made indoor and urban scenes are composed of a set of orthogonal and parallel *** robotics and computer vision,these scenes typically represented by the Manhattan-World *** accurate estimation of the Manhattan Frame,which consists of three orthogonal directions being used to represent the Manhattan-World,plays an important role in many applications,such as SLAM,scene understanding and 3 D *** this paper,a new method for accurately recovering the Manhattan frame from a single RGB-D image by using the orientation relevance is *** first extracts planes from the input single RGB-D *** three orthogonal dominant planes are determined by introducing the concept of orientation ***,the Manhattan Frame can be easily recovered from the obtained three orthogonal dominant *** with open dataset validate the proposed *** overall performance of the proposed method,which takes both accuracy and speed into account,is superior to that of the state-of-the-art *** is also applied on the application of scene annotation to confirm its applicability.

关键词： RGB-D data Indoor Scene Understanding Manhattan Frame Estimation

来源：评论

学校读者我要写书评

暂无评论

Adaptive planar vision marker composed of LED arrays for sensing under low visibility 5th

Adaptive planar vision marker composed of LED arrays for sen...

引用

5th International Conference on Robot Intelligence Technology and applications, RiTA 2017

作者： Kim, Kyukwang Hyun, Jieum Myung, Hyun Urban Robotics Laboratory Korea Advanced Institute of Science and Technology 291 Daehak-ro Daejeon34141 Korea Republic of

ISBN: (纸本)9783319784519

In image processing and robotic applications, two-dimensional (2D) black and white patterned planar markers are widely used. However, these markers are not detectable in low visibility environment and they are not changeable. This research proposes an active and adaptive marker node, which displays 2D marker patterns using light emitting diode (LED) arrays for easier recognition in the foggy or turbid underwater environments. Because each node is made to blink at a different frequency, active LED marker nodes were distinguishable from each other from a long distance without increasing the size of the marker. We expect that the proposed system can be used in various harsh conditions where the conventional marker systems are not applicable because of low visibility issues. The proposed system is still compatible with the conventional marker as the displayed patterns are identical. © Springer International Publishing AG, part of Springer Nature 2019.

关键词： Light emitting diodes

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：