检索结果-内蒙古大学图书馆

Fine-Grained Visual Classification via Progressive Multi-granularity Training of Jigsaw Patches 1

16th European Conference on Computer Vision, ECCV 2020

作者： Du, Ruoyi Chang, Dongliang Bhunia, Ayan Kumar Xie, Jiyang Ma, Zhanyu Song, Yi-Zhe Guo, Jun Pattern Recognition and Intelligent System Laboratory School of Artificial Intelligence Beijing University of Posts and Telecommunications Beijing China SketchX CVSSP University of Surrey Guildford United Kingdom

ISBN: (数字)9783030585655

ISBN: (纸本)9783030585648

Fine-grained visual classification (FGVC) is much more challenging than traditional classification tasks due to the inherently subtle intra-class object variations. Recent works are mainly part-driven (either explicitly or implicitly), with the assumption that fine-grained information naturally rests within the parts. In this paper, we take a different stance, and show that part operations are not strictly necessary – the key lies with encouraging the network to learn at different granularities and progressively fusing multi-granularity features together. In particular, we propose: (i) a progressive training strategy that effectively fuses features from different granularities, and (ii) a random jigsaw patch generator that encourages the network to learn features at specific granularities. We evaluate on several standard FGVC benchmark datasets, and show the proposed method consistently outperforms existing alternatives or delivers competitive results. The code is available at https://***/PRIS-CV/PMG-Progressive-Multi-Granularity-Training. © 2020, Springer Nature Switzerland AG.

关键词： Computer vision

来源：评论

学校读者我要写书评

暂无评论

A new RCNN-RF for efficient and robust parking lot available detection 16

A new RCNN-RF for efficient and robust parking lot available...

引用

16th International Conference on Digital Image Processing, ICDIP 2024

作者： Zhang, Qian Ran, Yaqin Luo, Di Xu, Kaili Shi, Jilaing School of Data Science and Information Engineering GuiZhou MinZu University Guiyang550025 China Guizhou Key Laboratory of Pattern Recognition and Intelligent System Guiyang550025 China

ISBN: (数字)9781510682917

ISBN: (纸本)9781510682900

It's very meaningful to conduct the driver to the parking space available in the parking lot clearly and accurately by computer vision and computational intelligence. While it is an extremely difficult task, because of the scene complexity in the parking lot. In this paper, a new computational intelligence technology is proposed to detect parking spaces available in the parking lot outdoor based on fusing of Region Convolution Neural Network (RCNN) and random forests (RF), and we called it as RCNN-RF. Within the RCNN-RF, a new activation function is design. And with the help of random forests, RCNN-RF overcomes scene complexity includes illumination, weather, and slight occlusion successfully. Extensive experimental results on the benchmark dataset named PKLot and the self-built dataset named GZMU-LOT outperform those of the state-of-the-art, and it shows the superiority of our proposed RCNN-RF method. © 2024 SPIE.

关键词： Random forests

来源：评论

学校读者我要写书评

暂无评论

BLOG EXTRACTION WITH TEMPLATEINDEPENDENT WRAPPER

BLOG EXTRACTION WITH TEMPLATEINDEPENDENT WRAPPER

引用

2010 2nd IEEE International Conference on Network Infrasturcture & Digital Content(2010年第二届IEEE网络基础设施与数字内容国际会议 IC-NIDC 2010)

作者： Zhixuan Zhang Chuang Zhang Zhiqing Lin Bo Xiao Pattern Recognition and Intelligent System Laboratory(PRIS) Beijing University of Posts and Telecomm Pattern Recognition and Intelligent System Laboratory(PRIS) Beijing University of Posts and Telecom

Rich information is contributed to blogs by millions of users all around the world with the development of blogsphere. However, few work has been done on the study of blog extraction so far. Unlike the traditional template-dependent wrapper, not only blog articles but also blogroll is extracted with template-independent wrapper in this paper. In our method, blog extraction is formalized as a machine learning problem and a template-independent wrapper is learned by using labeled blog pages from a single site. Testing pages are obtained from 10 popular Chinese blog sites. And experimental results on 300 real blog pages indicate that the proposed method can correctly extract data from blogs with the accuracy of 90％ or even above.

关键词： template-independent data extraction web mining

来源：评论

学校读者我要写书评

暂无评论

Quantile-based method for the recognition of handwritten Chinese characters

Quantile-based method for the recognition of handwritten Chi...

引用

International Conference on Signal Processing Proceedings (ICSP)

作者： Ren Junling Wang Chengquan Guo Jun Pattern Recognition and Intelligent System Laboratory BUPT Beijing China

In course of generating templates of handwritten Chinese characters, the method based on average of the samples features is popular. In this paper, an approach based on quantile is put forward to generate templates. According to our experiments, this method of template generating is robust than that one based on average and it improves the performance of recognition system in the case of not increasing any complexity of the algorithm. At the same time, the method is universal and can be in combination with diversified feature extraction methods.

关键词： Character recognition Handwriting recognition Character generation Robustness Statistics pattern recognition Feature extraction Gaussian distribution intelligent systems Utility programs

来源：评论

学校读者我要写书评

暂无评论

Research on Real-time Video Action Classification Based on Three-Dimensional Convolutional Neural Network

Research on Real-time Video Action Classification Based on T...

引用

IEEE International Conference on Network Infrastructure and Digital Content (IC-NIDC)

作者： Mingli Yu Chuang Zhang Ming Wu Pattern Recognition and Intelligent System Lab BUPT Beijing China

This paper describes a model for performing action classification in real-time video streaming. This model can simultaneously analyze the spatio-temporal information of video under the constraint of low delay. In addition, in order to prevent the model from judging motionless segments in the video as motion, the model in this article is equipped with the ability of distinguish the segments of motion from the stationary ones. The experimental results show that the model can complete the action classification task with little delay, which ensures that the classification result can be output in real time with the constant input of the video image.

关键词： Streaming media Optical flow Real-time systems Optical buffering Convolutional neural networks Image segmentation

来源：评论

学校读者我要写书评

暂无评论

A novel keypoint detection in wavelet pyramid space

International Journal of Digital Content Technology and its ...

引用

International Journal of Digital Content Technology and its Applications 2011年第6期5卷 55-62页

作者： Gao, Yanyan Zhang, Honggang Guo, Jun Pattern Recognition and Intelligent System Laboratory Beijing University of Posts and telecommunications Beijing China

Keypoint detection is important for object recognition, image retrieval, mosaicing etc., and has attracted ample research. In this paper, we propose a novel wavelet-based detector (NWBD) based on the previous researches on keypoint detection. NWBD is performed in wavelet pyramid space, it extracts the local extrema of the energy map computed by intra-scale coefficient product (ISCP) as the candidate keypoint, and then discards some points by Hessian matrix. In the experiments, the novel detector was compared with Harris detector and SIFT detector by the evaluation of repeatability, and it achieved better performance for some scenes in the database provided by Mikolajcyzk and Schmid, such as wall, trees, and graffiti.

关键词： Object recognition

来源：评论

学校读者我要写书评

暂无评论

Instance-level coupled subspace learning for fine-grained sketch-based image retrieval 14

Instance-level coupled subspace learning for fine-grained sk...

引用

Computer Vision - ECCV 2016 Workshops, Proceedings

作者： Xu, Peng Yin, Qiyue Qi, Yonggang Song, Yi-Zhe Ma, Zhanyu Wang, Liang Guo, Jun Pattern Recognition and Intelligent System Laboratory Beijing University of Posts and Telecommunications Beijing China National Lab of Pattern Recognition Institute of Automation Chinese Academy of Sciences Beijing China SketchX Lab School of Electronic Engineering and Computer Science Queen Mary University of London London United Kingdom

ISBN: (纸本)9783319466033

Fine-grained sketch-based image retrieval (FG-SBIR) is a newly emerged topic in computer vision. The problem is challenging because in addition to bridging the sketch-photo domain gap, it also asks for instance-level discrimination within object categories. Most prior approaches focused on feature engineering and fine-grained ranking, yet neglected an important and central problem: how to establish a finegrained cross-domain feature space to conduct retrieval. In this paper, for the first time we formulate a cross-domain framework specifically designed for the task of FG-SBIR that simultaneously conducts instancelevel retrieval and attribute prediction. Different to conventional phototext cross-domain frameworks that performs transfer on category-level data, our joint multi-view space uniquely learns from the instance-level pair-wise annotations of sketch and photo. More specifically, we propose a joint view selection and attribute subspace learning algorithm to learn domain projection matrices for photo and sketch, respectively. It follows that visual attributes can be extracted from such matrices through projection to build a coupled semantic space to conduct retrieval. Experimental results on two recently released fine-grained photo-sketch datasets show that the proposed method is able to perform at a level close to those of deep models, while removing the need for extensive manual annotations. © Springer International Publishing Switzerland 2016.

关键词： Learning algorithms

来源：评论

学校读者我要写书评

暂无评论

A proposed GM-GRNN model for the prediction of behavior in complex system

A proposed GM-GRNN model for the prediction of behavior in c...

引用

19th IASTED International Conference on Modelling and Simulation, MS 2008

作者： Pan, Wei Huang, Yupeng DeGaris, Hugo Pattern Recognition and Intelligent System Institute Information Science and Technology School Xiamen University China Artificial Intelligence Institute Information Science and Technology School Xiamen University China

ISBN: (纸本)9780889867420

This paper analyses the kernel of the General Regression Neural Network (GRNN) model in detail, and presents its deficiencies in the domain of complex systems forecasting. We import various aspects of the Broyden-Fletcher-Goldfarb- Shanno (BFGS) Quasi-Newton method and GM(1,h) algorithms to improve the kernel of the GRNN model. We then apply this modified model to the problem of unemployment forecasting in China, as an example of its ability to model time-varying environments.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Competing Ratio Loss for Multi-class image Classification 34

Competing Ratio Loss for Multi-class image Classification

引用

34th IEEE International Conference on Visual Communications and Image Processing, VCIP 2019

作者： Zhang, Ke Wang, Xinsheng Guo, Yurong Zhao, Zhenbing Ma, Zhanyu North China Electric Power University Department of Electronic and Communication Engineering Hebei China Pattern Recognition and Intelligent System Laboratory Beijing University of Posts and Telecommunications Beijing China

ISBN: (纸本)9781728137230

Cross-entropy loss function (CEL) is widely used for training a multi-class classification deep convolutional neural network (DCNN). While CEL has been successfully implemented in image classification tasks, it only focuses on the posterior probability of correct class when the labels of training images are one-hot. It cannot be discriminated against the classes not belong to correct class (wrong classes) directly. Negative Log Likelihood Ratio Loss (NLLR) is proposed to better discriminate the correct class from competing wrong classes. But optimization of the loss function is normally presented as a minimization problem. In training DCNN, the value of NLLR is not constantly positive or negative, which affects the convergence of NLLR adversely. So, we propose competing ratio loss (CRL), which calculates the posterior probability ratio between the correct class and competing wrong classes to better widen the difference between the probability of the correct class and the probabilities of wrong classes, which also assures the value of CRL is constantly positive. Through massive experiments, we demonstrate the effectiveness and robustness of CRL on deep convolutional neural networks, our CRL outperforms CEL and NLLR on CIFAR-10/100 datasets. © 2019 IEEE.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

intelligent Parking Big Data Cloud Platform Algorithm Research and Application 6

Intelligent Parking Big Data Cloud Platform Algorithm Resear...

引用

6th International Conference on Computer Information Science and Application Technology, CISAT 2023

作者： Yan, Xiaobo Zou, Xue Cai, Shan Wang, Lin Key Laboratory of Pattern Recognition and Intelligent System of Guizhou Province Guizhou Minzu University Guiyang China College of Date Science and Information Engineering Guizhou Minzu University Guiyang China

ISBN: (纸本)9781510668546

With the rapid development of the domestic economy and the increasing living standards of the people, the ownership of private cars has increased explosively. Currently, urban traffic congestion and parking difficulty have gradually become a hot topic of research from all walks of life. Parking difficulty is mainly reflected on the following aspects: first, the existing parking spaces in cities are not effectively utilized;The second reason is that drivers cannot grasp relevant information about parking spaces near the ground in real time. This article conducts in-depth research on the issue of parking difficulty. Based on the existing security monitoring systems in parking lots, real-time detection of parking spaces status is conducted based on computer vision, machine learning, and other methods. By building a comprehensive, synthesize, and multi-dimensional parking resources dynamic sensing system, a smart parking big data cloud platform is constructed to achieve the effect of intelligent indoor parking lots management. © 2023 SPIE.

关键词： Machine learning

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：