检索结果-内蒙古大学图书馆

Deep Learning for Object Detection:A Survey

computer Systems Science & engineering 2021年第8期38卷 165-182页

作者： Jun Wang Tingjuan Zhang Yong Cheng Najla Al-Nabhan Science and Technology Industry Department Nanjing University of Information Science&Technology210044China School of Computer&Software Jiangsu Engineering Center of Network MonitoringNanjing University of Information Science&Technology210044China Department Computer Science King Saud UniversityRiyadhKSA

Object detection is one of the most important and challenging branches of computer vision,which has been widely applied in people s life,such as monitoring security,autonomous driving and so on,with the purpose of locating instances of semantic objects of a certain *** the rapid development of deep learning algorithms for detection tasks,the performance of object detectors has been greatly *** order to understand the main development status of target detection,a comprehensive literature review of target detection and an overall discussion of the works closely related to it are presented in this *** paper various object detection methods,including one-stage and two-stage detectors,are systematically summarized,and the datasets and evaluation criteria used in object detection are *** addition,the development of object detection technology is ***,based on the understanding of the current development of target detection,we discuss the main research directions in the future.

关键词： Object detection convolutional neural network computer vision

来源：评论

学校读者我要写书评

暂无评论

Coverless image steganography based on SIFT and BOF

Coverless image steganography based on SIFT and BOF

引用

作者： Yuan, Chengsheng Xia, Zhihua Sun, Xingming School of Computer and Software Nanjing University of Information Science and Technology China Jiangsu Engineering Center of Network Monitoring Nanjing University of Information Science and Technology China

In order to realize secret communication, conventional steganography techniques can embed the secret message into carrier images by modifying the content and structure of carrier images. Although steganography techniques provide us with some more secure communication, the modification traces will be detected through using these brand-new steganalysis tools. Thereby, in order to eliminate the influence of modification traces, we propose a novel coverless image steganography scheme based on scale invarient feature transform and bag of feature. Different from conventional steganography, images have already contained the concealing information, so we can transmit natural images, whose features are the same with the secret information, to receivers. Firstly, the robust feature hash sequences are constructed with our protocol of feature. Then, the secret message is converted into bitstreams. Finally, these images with the seceret information are chosen and sent to receivers through using inverted index. Throughout the process, original images are never modified in our scheme, so our method can resist the analysis of existing steganalysis tools. Numerical experiments also indicate that our scheme has a desirable robustness to the common image attacks.

关键词： Steganography

来源：评论

学校读者我要写书评

暂无评论

Short text classification technology based on KNN+hierarchy SVM 12th

Short text classification technology based on KNN+hierarchy ...

引用

12th International Conference on Future Information Technology, FutureTech 2017

作者： Yin, Chunyong Shi, Lingfeng Wang, Jin School of Computer and Software Jiangsu Engineering Center of Network Monitoring Nanjing University of Information Science and Technology Nanjing China College of Information Engineering Yangzhou University Yangzhou China

ISBN: (纸本)9789811050404

A short text classification method based on combination of KNN and hierarchical SVM is proposed. First, the KNN algorithm is improved to get the K nearest neighbor class labels quickly, so as to effectively filter the candidate classes of documents. And then classify them from top to bottom using a multi-class sparse hierarchical SVM classifier. By this way, the document can be classified efficiently. © Springer Nature Singapore Pte Ltd. 2017.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

Improved data stream clustering algorithm for anomaly detection 12th

Improved data stream clustering algorithm for anomaly detect...

引用

12th International Conference on Future Information Technology, FutureTech 2017

作者： Yin, Chunyong Zhang, Sun Wang, Jin School of Computer and Software Jiangsu Engineering Center of Network Monitoring Nanjing University of Information Science and Technology Nanjing China College of Information Engineering Yangzhou University Yangzhou China

ISBN: (纸本)9789811050404

Intrusion detection provides important protection for network security and anomaly detection as a type of intrusion detection, can recognize the pattern of normal behaviors and label the behaviors which departure from normal pattern as abnormal behaviors. We think that the traditional methods based on dataset do not satisfy the needs of dynamic network environment. The network data stream is temporal and cannot be treated as static dataset. The concept and distribution of data objects is variety in different time stamps and the changing is unpredictable. Therefore, we propose an improved data stream clustering algorithm and design the frame of anomaly detection according to the improved algorithm. It can modify the established model with the changing of data stream and detect abnormal behaviors in time. © Springer Nature Singapore Pte Ltd. 2017.

关键词： Intrusion detection

来源：评论

学校读者我要写书评

暂无评论

Application of an improved data stream clustering algorithm in intrusion detection system 12th

Application of an improved data stream clustering algorithm ...

引用

12th International Conference on Future Information Technology, FutureTech 2017

作者： Yin, Chunyong Xia, Lian Wang, Jin School of Computer and Software Jiangsu Engineering Center of Network Monitoring Nanjing University of Information Science and Technology Nanjing China College of Information Engineering Yangzhou University Yangzhou China

ISBN: (纸本)9789811050404

With the continuous development of computer network technology, traditional intrusion detection system is short of good adaptability. Aiming at the traditional intrusion detection system is difficult to adapt to the increasing amount of data demand for real-time processing capability, this paper proposes a clustering algorithm based on sliding window data streams, based on which we build the IDS network security defense model. The experiment results show that the model is able to adapt to the high-speed network intrusion detection requirements. © Springer Nature Singapore Pte Ltd. 2017.

关键词： Data mining

来源：评论

学校读者我要写书评

暂无评论

A New Sequential Image Prediction Method Based on LSTM and DCGAN

引用

computers, Materials & Continua 2020年第7期64卷 217-231页

作者： Wei Fang Feihong Zhang Yewen Ding Jack Sheng School of Computer&Software Jiangsu Engineering Center of Network MonitoringNanjing University of Information Science&TechnologyNanjing210044China State Key Laboratory for Novel Software Technology Nanjing UniversityNanjing210023China Department of Economics FinanceInsurance and Risk Management University of Central ArkansasConway72035USA

Image recognition technology is an important field of artificial *** with the development of machine learning technology in recent years,it has great researches value and commercial *** a matter of fact,a single recognition function can no longer meet people’s needs,and accurate image prediction is the trend that people *** paper is based on Long Short-Term Memory(LSTM)and Deep Convolution Generative Adversarial networks(DCGAN),studies and implements a prediction model by using radar image *** adopt a stack cascading strategy in designing network connection which can control of parameter convergence *** new method enables effective learning of image features and makes predictive models to have greater generalization *** demonstrate that our network model is more robust and efficient in terms of timing prediction than 3DCNN and traditional *** sequential image prediction model architecture proposed in this paper is theoretically applicable to all sequential images.

关键词： Image prediction LSTM DCGAN

来源：评论

学校读者我要写书评

暂无评论

Non-Exchangeable Error Compensation for Strapdown Inertial Navigation System in High Dynamic Environment

引用

computers, Materials & Continua 2020年第8期64卷 973-986页

作者： Qi Wang Changsong Yang Shao’en Wu School of Computer and Software Nanjing University of Information Science and TechnologyNanjing210044China Jiangsu Engineering Center of Network Monitoring Nanjing University of Information Science and TechnologyNanjing210044China School of Automation Nanjing University of Information Science and TechnologyNanjing210044China Department of Computer Science Ball State UniversityMuncie47306USA

Strapdown non-exchangeable error compensation technology in high dynamic environment is one of the key technologies of strapdown inertial navigation *** platform is used in strapdown inertial navigation system instead of physical platform in traditional platform inertial navigation system,which improves reliability and reduces cost and volume of *** maximum error source of attitude matrix solution is the non-exchangeable error of rotation due to the non-exchangeable of finite rotation of rigid *** rotation non-exchangeable error reaches the maximum in coning motion,although it can be reduced by shortening the correction period and increasing the real-time *** equivalent rotation vector method is used to modify the attitude to reduce the coning error in this *** experiments show that the equivalent rotation vector method can effectively suppress the non-exchangeable error and improve the accuracy of attitude calculation.

关键词： Strapdown error compensation coning motion equivalent rotation vector simulation experiments

来源：评论

学校读者我要写书评

暂无评论

Combining active learning and semi-supervised for improving learning performance 11

Combining active learning and semi-supervised for improving ...

引用

4th International Symposium on Applied Sciences in Biomedical and Communication Technologies, ISABEL'11

作者： Ma, Tinghuai Ge, Jian Wang, Jin Jiangsu Engineering Center of Network Monitoring Nanjing University of Information Science and Technology Nanjing Jiangsu Province China School of Computer and Software Nanjing University of Information Science and Technology Nanjing Jiangsu Province China

ISBN: (纸本)9781450309134

In many learning tasks, there are abundant unlabeled samples but the number of labeled training samples is limited, because labeling the samples requires the efforts of human annotators and expertise. There are three major techniques for labeling the samples: semi-supervised learning, transductive learning and active learning. Semi-supervised and transductive learning deal with methods for automated exploiting unlabeled samples in addition to improve learning performance. Active learning deals with methods that assume the learner has control over the whole input space. So combing the advantage of semi-supervised learning and active learning is a practical technique for improving the learning performance. In this paper, a general framework of combing (Active Learning) AL and (Semi-Supervised Learning) SSL algorithms is proposed. Then the ensemble learning for combing AL and SSL algorithms is introduced, which is denoted by ASC (AL and SSL by Committee). At last, the ensemble learning and confidence measure of the ASC is discussed. © 2011 ACM.

关键词： Learning algorithms

来源：评论

学校读者我要写书评

暂无评论

Vehicle Re-Identication Model Based on Optimized DenseNet121 with Joint Loss

引用

computers, Materials & Continua 2021年第6期67卷 3933-3948页

作者： Xiaorui Zhang Xuan Chen Wei Sun Xiaozheng He Jiangsu Engineering Center of Network Monitoring Engineering Research Center of Digital ForensicsMinistry of EducationSchool of Computer and SoftwareNanjing University of Information Science&TechnologyNanjing210044China Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology Nanjing University of Information Science&TechnologyNanjing210044China Department of Civil and Environmental Engineering Rensselaer Polytechnic InstituteTroyNY12180USA

With the increasing application of surveillance cameras,vehicle re-identication(Re-ID)has attracted more attention in the eld of public *** Re-ID meets challenge attributable to the large intra-class differences caused by different views of vehicles in the traveling process and obvious inter-class similarities caused by similar *** existing methods focus on local attributes by marking local ***,these methods require additional annotations,resulting in complex algorithms and insufferable computation *** cope with these challenges,this paper proposes a vehicle Re-ID model based on optimized DenseNet121 with joint *** model applies the SE block to automatically obtain the importance of each channel feature and assign the corresponding weight to it,then features are transferred to the deep layer by adjusting the corresponding weights,which reduces the transmission of redundant information in the process of feature reuse in *** the same time,the proposed model leverages the complementary expression advantages of middle features of the CNN to enhance the feature expression ***,a joint loss with focal loss and triplet loss is proposed in vehicle Re-ID to enhance the model’s ability to discriminate difcult-to-separate samples by enlarging the weight of the difcult-to-separate samples during the training *** results on the VeRi-776 dataset show that mAP and Rank-1 reach 75.5%and 94.8%,***,Rank-1 on small,medium and large sub-datasets of Vehicle ID dataset reach 81.3%,78.9%,and 76.5%,respectively,which surpasses most existing vehicle Re-ID methods.

关键词： Vehicle re-identication densenet joint loss focal loss SE block

来源：评论

学校读者我要写书评

暂无评论

A novel convolution neural network for background segmentation recognition 4th

A novel convolution neural network for background segmentati...

引用

4th International Conference on Cloud Computing and Security, ICCCS 2018

作者： Fang, Wei Ding, Yewen Zhang, Feihong School of Computer and Software Jiangsu Engineering Center of Network Monitoring Nanjing University of Information Science and Technology Nanjing China State Key Laboratory for Novel Software Technology Nanjing University Nanjing China

ISBN: (纸本)9783030000059

The convolution neural network for image classification is an application of deep learning on image processing. Convolutional neural networks have the advantage of being able to convolve directly with image pixels and extract image features from image pixels. This approach is closer to the treatment of the human brain’s visual system. However, up to now, it is impossible to achieve 100% classification accuracy regardless of any kind of convolutional neural network models. At the same time, we also found that sometimes the picture background will also affect the recognition effect of the neural network on the picture, and after removing the picture background, it can correctly recognize the same picture. Therefore, we design a model based on the convolutional neural network to remove the image background for some specific images (such as selfies, animals, flowers, etc.) here, and then the image after processing will be identified and classified. Experiments show that the proposed method can maintain a high level of integrity of the target to be detected in the image after removing the background (here the indicator is Intersection Over Union, IOU);moreover, through multiple classification models verify that the classification accuracy of some background-removed pictures is significantly higher than that of pictures without any treatment. © Springer Nature Switzerland AG 2018.

关键词： Image classification

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：