检索结果-内蒙古大学图书馆

1st IEEE international conference on Artificial Intelligence and Knowledge Engineering (AIKE)

作者： Ngoc-Quang Nguyen Lee, Sang-Woong Gachon Univ Pattern Recognit & Machine Learning Lab Seongnam South Korea

ISBN: (纸本)9781538695555

Colorectal cancer is the third most common cancer which causes of cancer-related deaths. Therefore, early diagnosis of polyps by colonoscopy could result in successful treatment. Diagnosis of polyps in colonoscopy videos is a challenging task due to variations in the size and shape of polyps. In this paper, we propose a polyp segmentation method based on the encoder decoder network. Performance of the method is enhanced by two strategies, we perform a novel database augmentation method for colonoscopy images in the training phase. Besides, in the test phase, we perform an effective prediction by combining multi model to compare the probability of each image that is produced by the network. Evaluation of the proposed method using the ETIS-LariPolypDB [9] database shows that our proposed method outperforms state-of-the-art results.

关键词： colorectal segmentation CRC encoder-decoder multi-model

来源：评论

学校读者我要写书评

暂无评论

An Ensemble Model for Error Modeling with Pseudoinverse learning Algorithm

An Ensemble Model for Error Modeling with Pseudoinverse Lear...

引用

IEEE international conference on Systems, Man and Cybernetics

作者： Sibo Feng Xiaodan Deng Ping Guo Bo Zhao Qian Yin Hongfeng Wang Image Processing and Pattern Recognition Laboratory Beijing Normal University Beijing China School of Information Management Dezhou University Shandong China

In Bayesian theory, the maximum posterior estimator uses prior information to estimate the noise in the machine learning model by adding the regularization term. The regularization terms L 1 and L 2 correspond to Laplacian prior and Guassian prior, respectively. In existing deep learning models, in order to use the gradient descent optimization algorithm and achieve good results, most models take L 2 regularization as the regularization term of the network model to fit the complex Guassian noise. However in practice, the Laplace noise and the Guassian noise are both considered as data noise. For multi-layer perceptrons, the difficulty caused by adding L 1 and L 2 into the optimization function of the network is solved by proposing an ensemble model for error modeling through adopting the divide and conquer strategy. First, several base learners are trained to fit different noise distributions of data, then the final results can be obtained by taking the results of each base leaner as new data to train a meta leaner, and get the final results. Among them, coordinate regression method is used to solve L 1 loss, while the pseudo-inverse learning algorithm is employed to solve L 2 loss. Both methods are nongradient optimization algorithms. The comparison results of the model on several data sets show that the proposed ensemble model achieves better performance.

关键词： Training Optimization Data models Neural networks Computational modeling Matrix decomposition Estimation

来源：评论

学校读者我要写书评

暂无评论

A New Real-Time FPGA-Based Implementation of K-Means Clustering for images 1st

A New Real-Time FPGA-Based Implementation of K-Means Cluster...

引用

1st international conference on Intelligent Manufacturing and Internet of Things (IMIOT) / 5th international conference on Computing for Sustainable Energy and Environment (ICSEE)

作者： Deng, Tiantai Crookes, Danny Siddiqui, Fahad Woods, Roger Queens Univ Belfast Univ Rd Belfast Antrim North Ireland

ISBN: (纸本)9789811323843;9789811323836

As an unsupervised machine-learning algorithm, K-means clustering for images has been widely used in image segmentation. The standard Lloyd's algorithm iteratively allocates all image pixels to clusters until convergence. The processing requirement can be a problem for high-resolution images and/or real-time systems. In this paper, we present a new histogram-based algorithm for K-means clustering, and its FPGA implementation. Once the histogram has been constructed, the algorithm is O(GL) for each iteration, where GL is the number of grey levels. On a Xilinx ZedBoard, our algorithm achieves 140 FPS (640 x 480 images, running at 150 MHz, 4 clusters, 25 iterations), including final image reconstruction. At 100 MHz, it achieves 95 FPS. It is 7.6 times faster than the standard Lloyd's algorithm, but uses only approximately half of the resources, while giving the same results. The more iterations, the bigger the speed-up. For 50 iterations, our algorithm is 10.2 times faster than the Lloyd's approach. Thus for all cases our algorithm achieves real time performance whereas Lloyd's struggles to do so. The number of clusters (up to a user-defined limit) and the initialization method (one of three) can be selected at runtime.

关键词： Unsupervised machine learning Data processing K-means clustering FPGA acceleration

来源：评论

学校读者我要写书评

暂无评论

Patient-Specific Modeling for pattern Identification in Clinical Events Using Data Mining Models and Evolutionary Computing 1st

引用

1st international conference on Soft Computing: Theories and Applications, SoCTA 2016

作者： Yadav, Sourabh Saroliya, Anil Amity School of Engineering and Technology Amity University JaipurRajasthan India

ISBN: (纸本)9789811056864

Nowadays due to the technological advancement in collecting, processing and producing the information available, it is quite challenging to obtain the relevant information out from various sources of data. According to earlier research in homeopathy, a pattern system of treatment is being followed, which ultimately helps in finding the cure for the disease. One of the main goals of this system of treatment is to individualize the model in a manner that patient-specific prediction can be fixated with the help of pattern identification. By adopting the context of homeopathy system, the aim of this paper focuses on identifying appropriate parameters from a large quantity of raw clinical data, which are responsible for developing celiac disease in the human body. The result is based on the extensive quantitative research of a clinical database, shows that one parameter is correlating with another and such kind of pattern identification will be useful in the curing of celiac disease. © 2018, Springer Nature Singapore Pte Ltd.

关键词： machine learning

来源：评论

学校读者我要写书评

暂无评论

Generation of High Resolution Medical images Using Super Resolution via Sparse Representation 3rd

Generation of High Resolution Medical Images Using Super Res...

引用

3rd international Afro-European conference for Industrial Advancement (AECIA)

作者： Asif, Muhammad Khan, Shoab Ahmed Hassan, Taimur Akram, M. Usman Shaukat, Arslan Natl Univ Sci & Technol NUST H-12 Islamabad Pakistan

ISBN: (纸本)9783319608341;9783319608334

Generation of high resolution (HR) of images is very important in image processing and computer vision applications such as: CT scan-chest, MRI, X-rays and Optical Coherence Tomography (OCT). In medical imaging field images are very frequently using for analyzing different type of disease symptoms. HR medical images are required for proper diagnosing but due to the hardware limitation it is very difficult. In this paper we propose a HR image generation technique based on image statistic research for single low resolution (LR) input image-using joint dictionary learning. The view of LR image is like down sampled version of HR image and its patches are considered to have a sparse representation with respect to an over complete dictionary. Sparse representation can be recovered under mild condition with compressed sensing perspective. This study explores the use of super resolution (SR), due to learning there is no need to aligned subpixels of different LR images. For validity of our research, we use this technique on OCT and Lungs images and train three dictionaries (i) using OCT images (U) using Lungs images and (iii) using multiple different natural (MDN)-HR images. images are produced with using MDN-HR and for both medical images and also used OCT images dictionary for OCT images and lungs images dictionary for lungs images. We compare our result with previous SR proposed technique-from all aspects dictionary learning approach have superiority. Proposed dictionary learning based technique produced enhanced and upsampled image.

关键词： High resolution Super resolution Sparse representation CT scan of chest Lungs Optical Coherence Tomography Dictionary learning

来源：评论

学校读者我要写书评

暂无评论

Foreword to the Special Issue on Analysis of Multitemporal Remote Sensing images

引用

IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING 2018年第12期11卷 4548-4550页

作者： Bruzzone, L. Deronde, B. Swinnen, E. Univ Trento Dept Comp Sci & Informat Engn I-38123 Trento Italy VITO Remote Sensing B-2400 Mol Belgium

The thirteen papers in this special section were presented at the 9th international Workshop on the Analysis of Multitemporal Remote Sensing images (MultiTemp 2017), hosted by VITO Remote Sensing on June 27-29, 2017.

关键词： Special issues and sections Meetings Vegetation mapping Time series analysis Unsupervised learning Change detection algorithms Deep learning

来源：评论

学校读者我要写书评

暂无评论

1st international conference on Smart Multimedia, ICSM 2018

1st International Conference on Smart Multimedia, ICSM 2018

引用

1st international conference on Smart Multimedia, ICSM 2018

ISBN: (纸本)9783030043742

The proceedings contain 39 papers. The special focus in this conference is on Smart Multimedia. The topics include: Towards maritime videosurveillance using 4K videos;A heterogeneous image fusion algorithm based on LLC coding;Towards the identification of Parkinson’s disease using only T1 MR images;Atlas-free method of periventricular hemorrhage detection from preterm infants’ T1 MR images;when smart signal processing meets smart imaging;a regularized nonnegative third order tensor decomposition using a primal-dual projected gradient algorithm: Application to 3D fluorescence spectroscopy;Adaptive dithering using curved Markov-Gaussian noise in the quantized domain for mapping SDR to HDR image;a flexible method for time-of-flight camera calibration using random forest;a survey on vision-based hand gesture recognition;affectional ontology and multimedia dataset for sentiment analysis;research on path planning method of an unmanned vehicle in urban road environments;detecting attention in pivotal response treatment video probes;person authentication by air-writing using 3D sensor and time order stroke context;synthetic vision assisted real-time runway detection for infrared aerial images;detection-based online multi-target tracking via adaptive subspace learning;a deep learning approach to predict crowd behavior based on emotion;object tracking in hyperspectral videos with convolutional features and kernelized correlation filter;learning 3DMM deformation coefficients for rendering realistic expression images;semi-supervised adversarial image-to-image translation;a step beyond generative multi-adversarial networks;predicting student seating distribution based on social affinity;adversarial training for dual-stage image denoising enhanced with feature matching;IVUS-Net: An intravascular ultrasound segmentation network;a simplified active calibration algorithm for focal length estimation.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Face recognition Technology based on Kalman Filter

Face Recognition Technology based on Kalman Filter

引用

2019 9th international conference on Management and Computer Science(ICMCS 2019)

作者： Lili Wang School of Information Engineering Harbin University

In this paper, a combination of basic Haarlike features proposed by Viola-Jones an extended set of Haarlike features proposed by Lienhart et al will be researched and implemented in the face detection phase using matlab and a set of proposed rules for Adaptive Boosting(Adaboost). Followed by a method of Lip Localization using YCbCr colourspace, lip contour estimation with the Van-Chese method of active contours without edges and tracking using kalman filter algorithm. The significance of the item is that it can be applied to human-computer interaction, security and surveillance, video communication and compression, medical imaging, video editing and robot vision applications. This paper can also serve as groundwork for future research into automatic lip reading systems which may allow it possible for an application to know what one is saying in a video simply by reading their lips. Introduction Detection as the first step is necessary for any image processing methods that require working with the face [1,2]. In the past 20 years, research in the fields of computer vision, pattern recognition and machine learning has attracted researchers all over the world. Face recognition and other video coding techniques have recently received a great amount of attention from researchers. Various research methods have been proposed and implemented particularly in the areas of digital image analysis which spans especially through face detection, recognition and other related areas [3]. Most of these proposed methods assume during experiments that the images are in frontal views and under controlled conditions such as background and lighting which are not true in a real case scenario. For instance in the case of this item, after face detection mouth localization and must be done so that further estimation of the lip contours can be easily achievable. Different method of facial features have also been proposed and tested and shall be discussed in the sections that follow [4,5,6]. Vi

关键词： Lip movement Face recognition pattern recognition Computer vision Human computer interaction

来源：评论

学校读者我要写书评

暂无评论

1st international conference on Smart System, Innovations and Computing, SSIC 2017

1st International Conference on Smart System, Innovations an...

引用

1st international conference on Smart System, Innovations and Computing, SSIC 2017

ISBN: (纸本)9789811058271

The proceedings contain 84 papers. The special focus in this conference is on Smart System, Innovations and Computing. The topics include: Robust and efficient routing in three-dimensional underwater sensor network;limiting route request flooding using velocity constraint in multipath routing protocol;forensic writer identification with projection profile representation of graphemes;secure public auditing using batch processing for cloud data storage;workflow scheduling in cloud computing environment using bat algorithm;recognition of telecom customer’s behavior as data product in CRM big data environment;a perspective analysis of phonological structure in Indian sign language;air pollution prediction using extreme learning machine: A case study on Delhi (India);Load-balanced energy-enhanced routing protocol for clustered Bee-Ad Hoc MANETs;ioT-based solution for food adulteration;A novel self-transforming image encryption algorithm using intrinsically mutating PRNG;A model of computation of QoS of WSCDL roles from traces of competing web services;a model to measure readability of captions with temporal dimension;readability assessment-cum-evaluation of government department websites of Rajasthan;a novel technique for frequently searching items with alarm method;a novel system of automated wheel bed with android phone and Arduino;a secure file transfer using the concept of dynamic random key, transaction id and validation key with symmetric key encryption algorithm;smart control based energy management setup for space heating;HAAS: Intelligent cloud for smart health care solutions;Compact slotted microstrip patch antenna with multiband characteristics for WLAN/WiMAX;SEE THROUGH approach for the solution to node mobility issue in underwater sensor network (UWSN);arduino controlled chessboard.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Graph Regularized Discriminative Joint Concept Factorization for Data Representation 1st

Graph Regularized Discriminative Joint Concept Factorization...

引用

1st CCF international conference on Artificial Intelligence (CCF-ICAI)

作者： Long, Xianzhong Cheng, Cheng Nanjing Univ Posts & Telecommun Sch Comp Sci & Technol Nanjing 210023 Jiangsu Peoples R China Nanjing Univ Aeronaut & Astronaut Coll Comp Sci & Technol Nanjing 211106 Jiangsu Peoples R China Jiangsu Key Lab Big Data Secur & Intelligent Proc Nanjing 210023 Jiangsu Peoples R China

ISBN: (纸本)9789811321221;9789811321214

In the computer vision and pattern recognition fields, how to effectively represent the data is very important. A good data representation scheme can improve subsequent classification or clustering tasks. In this paper, a graph regularized discriminative joint concept factorization (GDJCF) method is proposed for data representation. In the GDJCF, we make use of three aspects of information, including multi-features joint learning, local manifold geometry structure and sample label. The learned representations are expected to have discriminative ability and better describe the raw data. We provide the corresponding multiplicative update solutions for the optimization framework, together with the convergence verification. Clustering results on five image data sets show that the GDJCF outperforms the state-of-the-art algorithms in terms of accuracy and normalized mutual information.

关键词： Joint concept factorization Multi-features learning Graph regularized Clustering

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：