检索结果-内蒙古大学图书馆

18th Iberoamerican Congress on pattern Recognition, CIARP 2013

作者： San Biagio, Marco Martelli, Samuele Crocco, Marco Cristani, Marco Murino, Vittorio Istituto Italiano di Tecnologia Pattern Analysis and Computer Vision Via Morego 30 16163 Genova Italy Università degli Studi di Verona Dipartimento di Informatica Strada le Grazie 15 37134 Verona Italy

ISBN: (纸本)9783642418211

Encoding an object essence in terms of self-similarities between its parts is becoming a popular strategy in computer vision. In this paper, a new similarity-based descriptor, dubbed Structural Similarity Cross-Covariance Tensor is proposed, aimed to encode relations among different regions of an image in terms of cross-covariance matrices. The latter are calculated between low-level feature vectors extracted from pairs of regions. The new descriptor retains the advantages of the widely used covariance matrix descriptors [1], extending their expressiveness from local similarities inside a region to structural similarities across multiple regions. The new descriptor, applied on top of HOG, is tested on object and scene classification tasks with three datasets. The proposed method always outclasses baseline HOG and yields significant improvement over a recently proposed self-similarity descriptor in the two most challenging datasets. © Springer-Verlag 2013.

关键词： Covariance matrix

来源：评论

学校读者我要写书评

暂无评论

Designing 3-D nonlinear diffusion filters for high performance cluster computing

Designing 3-D nonlinear diffusion filters for high performan...

引用

24th German Association for pattern Recognition Symposium, DAGM 2002

作者： Bruhn, Andrés Jakob, Tobias Fischer, Markus Kohlberger, Timo Weickert, Joachim Brüning, Ulrich Schnörr, Christoph Mathematical Image Analysis Group Mathematics and Computer Science Saarland University Building 27.1 Saarbrücken66041 Germany Computer Architecture Group Department of Mathematics and Computer Science University of Mannheim Mannheim68131 Germany Computer Vision Graphics and Pattern Recognition Group Department of Mathematics and Computer Science University of Mannheim Mannheim68131 Germany

ISBN: (纸本)354044209X

This paper deals with parallelization and implementation aspects of PDE based image processing models for large cluster environments with distributed memory. As an example we focus on nonlinear isotropic diffusion filtering which we discretize by means of an additive operator splitting (AOS). We start by decomposing the algorithm into small modules that shall be parallelized separately. For this purpose image partitioning strategies are discussed and their impact on the communication pattern and volume is analyzed. Based on the results we develop an algorithmic implementation with excellent scaling properties on massively connected low latency networks. Test runs on a high–end Myrinet cluster yield almost linear speedup factors up to 209 for 256 processors. This results in typical denoising times of 0.5 seconds for five iterations on a 256 × 256 × 128 data cube. © Springer-Verlag Berlin Heidelberg 2002.

关键词： Cluster computing

来源：评论

学校读者我要写书评

暂无评论

Adversarial feature augmentation for unsupervised domain adaptation

arXiv

引用

arXiv 2017年

作者： Volpi, Riccardo Morerio, Pietro Savarese, Silvio Murino, Vittorio Pattern Analysis & Computer Vision Istituto Italiano di Tecnologia Stanford Vision and Learning Lab Stanford University Computer Science Department Università di Verona

Recent works showed that Generative Adversarial Networks (GANs) can be successfully applied in unsupervised domain adaptation, where, given a labeled source dataset and an unlabeled target dataset, the goal is to train powerful classifiers for the target samples. In particular, it was shown that a GAN objective function can be used to learn target features indistinguishable from the source ones. In this work, we extend this framework by (i) forcing the learned feature extractor to be domain-invariant, and (ii) training it through data augmentation in the feature space, namely performing feature augmentation. While data augmentation in the image space is a well established technique in deep learning, feature augmentation has not yet received the same level of attention. We accomplish it by means of a feature generator trained by playing the GAN minimax game against source features. Results show that both enforcing domain-invariance and performing feature augmentation lead to superior or comparable performance to state-of-the-art results in several unsupervised domain adaptation benchmarks. Copyright © 2017, The Authors. All rights reserved.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Functionally Classifying Genes from Microarray Data Using Linear and Non-linear Data Projection

Functionally Classifying Genes from Microarray Data Using Li...

引用

ACS/IEEE International Conference on computer Systems and Applications

作者： J. Shaik M. Yeasin Computer Vision Pattern and Image Analysis (CVPIA) laboratory Department of Electrical and Computer Engineering University Of Memphis Memphis TN USA

来源：评论

学校读者我要写书评

暂无评论

Clustered subsampling for clinically informed diagnostic brain mapping

Clustered subsampling for clinically informed diagnostic bra...

引用

15th International Conference on Information Fusion, FUSION 2012

作者： Björnsdotter, Malin Sona, Diego Rosenthal, Sonny Dauwels, Justin School of Electrical and Electronics Engineering Nanyang Technological University Singapore Singapore Pattern Analysis and Computer Vision Fondazione Istituto Italiano di Tecnologia Italy Wee Kim Wee School of Communication and Information Nanyang Technological University Singapore Singapore

ISBN: (纸本)9780982443859

Brain based diagnostic systems have recently received attention as a tool in the characterization and diagnosis of a variety of neurodevelopmental and psychiatric disorders. Nonetheless, a majority of disorders are still diagnosed entirely based on symptom assessments and behavioral correlates. We therefore propose a method for fusing brain responses with clinical measures for improved diagnosis. To this end, we utilized the flexibility of clustered random subspace brain mapping to detect regions where brain responses in conjunction with a clinical measure could reliably differentiate patients from control subjects. We demonstrate the approach on realistically simulated functional magnetic resonance imaging (fMRI) brain activity and a clinical parameter. We show that the method efficiently identifies brain regions where fused analysis of brain responses and clinical parameters improves diagnosis compared to either measure alone. The proposed method is easy to implement and highly flexible, offering an appealing basis for multimodal brain mapping. © 2012 ISIF (Intl Society of Information Fusi).

关键词： Brain mapping

来源：评论

学校读者我要写书评

暂无评论

Robust object tracking in crowd dynamic scenes using explicit stereo depth

Robust object tracking in crowd dynamic scenes using explici...

引用

11th Asian Conference on computer vision, ACCV 2012

作者： Li, Chi Lu, Le Hager, Gregory D. Tang, Jianyu Wang, Hanzi Cognitive Science Department Xiamen University Xiamen China Imaging and Computer Vision Siemens Corporate Research Princeton NJ United States Computer Science Department Johns Hopkins University Baltimore MD United States Center for Pattern Analysis and Machine Intelligence Xiamen University China

ISBN: (纸本)9783642374302

In this paper, we exploit robust depth information with simple color-shape appearance model on single object tracking in crowd dynamic scenes. Since binocular video streams are captured from a moving camera rig, background subtraction cannot provide a reliable enhancement of region of interest. Our main contribution is a novel tracking strategy to employ explicit stereo depth to track and segment object in crowd dynamic scenes with occlusion handling. Appearance cues including color and shape play a secondary role to further extract the foreground acquired by previous depth-based segmentation. The proposed depth-driven tracking approach can largely alleviate the drifting issue, especially when the object frequently interacts with similar background in long sequence tracking. The problems caused by rapid object appearance change can also be avoided due to the stability of the depth cue. Furthermore, we propose a new, yet simple and effective depth-based scheme to cope with complete occlusion in tracking. From experiments on a large collection of challenging outdoor and indoor sequences, our algorithm demonstrates accurate and reliable tracking performance which outperforms other state-of-the-art competing algorithms. © 2013 Springer-Verlag.

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

PERSON RE-IDENTIFICATION WITH A PTZ CAMERA: AN INTRODUCTORY STUDY

PERSON RE-IDENTIFICATION WITH A PTZ CAMERA: AN INTRODUCTORY ...

引用

IEEE International Conference on Image Processing

作者： Pietro Salvagnini Loris Bazzani Marco Cristani Vittorio Murino Istituto Italiano di Tecnologia Pattern Analysis & Computer Vision Via Morego 30 16163 Genova - Italy

ISBN: (纸本)9781479923427

We present an introductory study that paves the way for a new kind of person re-identification, by exploiting a single Pan-Tilt-Zoom (PTZ) camera. PTZ devices allow to zoom on body regions, acquiring discriminative visual patterns that enrich the appearance description of an individual. This intuition has been translated into a statistical direct re-identification scheme, which collects two images for each probe subject: the first image captures the probe individual, focusing on the whole body;the second can be a zoomed body part (head, torso or legs) or another whole body image, and is the outcome of an action-selection mechanism, driven by feature selection principles. The validation of this technique is also explored: in order to allow repeatability, two novel multi-resolution benchmarks have been created. On these data, we demonstrate that our approach selects effective actions, by focusing on body portions which discriminate each subject. Moreover, we show that the proposed compound of two images overwhelms standard multi-shot descriptions, composed by many more pictures.

关键词： Person Re-identification Pan-Tilt-Zoom camera Cameras Body Regions Probes Body Image Leg images Whole Body Imaging Body Parts

来源：评论

学校读者我要写书评

暂无评论

Online Evaluation of Textile Yarn Evenness Using Fast Exposure Imaging and Image Processing Techniques

Online Evaluation of Textile Yarn Evenness Using Fast Exposu...

引用

Automation, Control and Robots (ICACR), International Conference on

作者： Noman Haleem Matteo Bustreo Alessio Del Bue University of Groningen Leeuwarden The Netherlands Pattern Analysis & Computer Vision Istituto Italiano di Tecnologia Genova Italy

Evenness is regarded as the most important attribute of textile yarn quality as irregular yarns reduce efficiency of spinning process and deteriorate quality of resultant fabrics. In traditional yarn spinning process, evenness is tested offline using laboratory scale testers through sampling from a whole lot of yarns. The lag associated with this approach limits agility in responding to a variety of process variations whereas limited sampling restricts achieving fine grained yarn quality insights. These factors together result in limited control on yarn spinning process and eventually compromise quality of the yarn *** aim to address this problem by developing a new yarn evenness testing system, which can be deployed on an individual spinning position to evaluate yarn quality in an online manner. The proposed system, which is based on a combination of fast exposure imaging and image processing techniques, was first validated using a control set of yarn specimens and later tested in normal yarn production through comparison with an existing offline yarn evenness tester. The results demonstrate promising potential of the proposed system for wider application in yarn spinning industry.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A multiple kernel learning approach to multi-modal pedestrian classification

A multiple kernel learning approach to multi-modal pedestria...

引用

International Conference on pattern Recognition

作者： Marco San-Biagio Aydın Ulaş Marco Crocco Marco Cristani Umberto Castellani Vittorio Murino Pattern Analysis & Computer Vision Istituto Italiano di Tecnologia Genova Italy Department of Computer Science University of Verona Verona Italy

ISBN: (纸本)9781467322164

Pedestrian detection is a key problem in many computer vision applications, especially in surveillance and security systems. To this end, information integration from different imaging modalities, such as thermal infrared and visible spectrum, can significantly improve the detection rate in respect to mono-modal strategies. For this reason, an effective fusion scheme is necessary to combine the information presented by multiple sensors. In this paper, we propose a pedestrian classification method based on the multiple kernel learning framework; standard pixel features (such as spatial derivatives) from both imaging modalities are employed to learn several feature-related basic kernels and a compound kernel is found as an optimized linear combination of basic kernels. Finally the compound kernel is used to train an SVM. Experiments performed on the OTCBVS dataset [1], demonstrate that our recipe definitely outclasses a wide set of literature fusion modalities.

关键词： Kernel Feature extraction Vectors Training Imaging Support vector machines Standards

来源：评论

学校读者我要写书评

暂无评论

Fully-automated analysis of muscle fiber images with combined region and edge-based active contours

Fully-automated analysis of muscle fiber images with combine...

引用

Workshops Bildverarbeitung fur die Medizin: Algorithmen Systeme Anwendungen, BVM 2006 - Workshop on Image Processing for Medicine: Algorithms - Systems - Applications, BVM 2006

作者： Brox, Thomas Kim, Yoo-Jin Weickert, Joachim Feiden, Wolfgang Computer Vision and Pattern Recognition Group University of Bonn Romerstr. 164 53117 Bonn Germany Institute of Neuropathology Saarland University 66421 Homburg/Saar Germany Math. Image Analysis Group Saarland University 66041 Saarbrücken Germany

ISBN: (纸本)9783540321361

This paper presents a new approach to automated muscle fiber analysis based on segmenting myofibers with combined region and edge-based active contours. It provides reliable and fully-automated processing, thus, enabling time-saving batch processing of the entire biopsy sample stemming from routinely HE-stained cryostat sections. The method combines color, texture, and edge cues in a level set based active contour model succeeded by a refinement with morphological filters. Falsepositive segmentations as compared to former methods are minimized. A quantitative comparison between manual and automated analysis of muscle fibers images did not reveal any significant differences. We gratefully acknowledge partial funding by the DFG.

关键词： Batch data processing

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：