检索结果-南通市图书馆

Real-Time Facial Pose Estimation and Tracking by Coarse-to-Fine Iterative Optimization

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Tsinghua Science and Technology 2020年第5期25卷 690-700页

作者： Xiaolong yang Xiaohong Jia Mengke Yuan dong-ming yan the Key Laboratory of Mathematics Mechanization Academy of Mathematics and Systems SciencesChinese Academy of SciencesBeijing 100190China the University of Chinese Academy of Sciences Beijing 100049China the National Laboratory of Pattern Recognition Institute of AutomationChinese Academy of SciencesBeijing 100190China

We present a novel and efficient method for real-time multiple facial poses estimation and tracking in a single frame or ***,we combine two standard convolutional neural network models for face detection and mean shape learning to generate initial estimations of alignment and ***,we design a bi-objective optimization strategy to iteratively refine the obtained *** strategy achieves faster speed and more accurate ***,we further apply algebraic filtering processing,including Gaussian filter for background removal and extended Kalman filter for target prediction,to maintain real-time tracking *** general RGB photos or videos are required,which are captured by a commodity monocular camera without any priori or *** demonstrate the advantages of our approach by comparing it with the most recent work in terms of performance and accuracy.

关键词： facial pose recognition facial pose estimation real-time tracking

Learning local shape descriptors for computing non-rigid dense correspondence

在线全文

学校读者我要写书评

暂无评论

Computational Visual Media 2020年第1期6卷 95-112页

作者： Jianwei Guo Hanyu Wang Zhanglin Cheng Xiaopeng Zhang dong-ming yan National Laboratory of Pattern Recognition Institute of AutomationChinese Academy of SciencesBeijing 100190China University of Maryland-College Park MarylandUSA Shenzhen VisuCA Key Lab Shenzhen Institutes of Advanced TechnologyChinese Academy of SciencesShenzhenChina

A discriminative local shape descriptor plays an important role in various *** this paper,we present a novel deep learning framework that derives discriminative local descriptors for deformable 3D *** use local"geometry images"to encode the multi-scale local features of a point,via an intrinsic parameterization method based on geodesic polar *** new parameterization provides robust geometry images even for badly-shaped triangular *** a triplet network with shared architecture and parameters is used to perform deep metric learning;its aim is to distinguish between similar and dissimilar pairs of ***,a newly designed triplet loss function is minimized for improved,accurate training of the triplet *** solve the dense correspondence problem,an efficient sampling approach is utilized to achieve a good compromise between training performance and descriptor *** testing,given a geometry image of a point of interest,our network outputs a discriminative local descriptor for *** testing of non-rigid dense shape matching on a variety of benchmarks demonstrates the superiority of the proposed descriptors over the state-of-the-art alternatives.

关键词： local feature descriptor triplet CNN dense correspondence geometry image non-rigid shape

Distinguishing Computer-Generated Images from Natural Images Using Channel and Pixel Correlation

在线全文

学校读者我要写书评

暂无评论

Journal of Computer Science & Technology 2020年第3期35卷 592-602页

作者： Rui-Song Zhang Wei-Ze Quan Lu-Bin Fan Li-ming Hu dong-ming yan National Laboratory of Pattern Recognition Institute of AutomationChinese Academy of SciencesBeijing 100190China School of Artificial Intelligence University of Chinese Academy of SciencesBeijing 100049China Alibaba Group Hangzhou 310023China State Key Laboratory of Hydro-Science and Engineering Tsinghua UniversityBeijing 100084China

With the recent tremendous advances of computer graphics rendering and image editing technologies,computergenerated fake images,which in general do not reflect what happens in the reality,can now easily deceive the inspection of human visual *** this work,we propose a convolutional neural network(CNN)-based model to distinguish computergenerated(CG)images from natural images(NIs)with channel and pixel *** key component of the proposed CNN architecture is a self-coding module that takes the color images as input to extract the correlation between color channels *** previous approaches that directly apply CNN to solve this problem,we consider the generality of the network(or subnetwork),i.e.,the newly introduced hybrid correlation module can be directly combined with existing CNN models for enhancing the discrimination capacity of original *** results demonstrate that the proposed network outperforms state-of-the-art methods in terms of classification *** also show that the newly introduced hybrid correlation module can improve the classification accuracy of different CNN architectures.

关键词： natural image computer-generated image channel and pixel correlation convolutional neural network

Joint specular highlight detection and removal in single images via Unet-Transformer

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computational Visual Media 2023年第1期9卷 141-154页

作者： Zhongqi Wu Jianwei Guo Chuanqing Zhuang Jun Xiao dong-ming yan Xiaopeng Zhang National Laboratory of Pattern Recognition Institute of AutomationChinese Academy of SciencesBeijing 100190China The School of Artificial Intelligence University of Chinese Academy of SciencesBeijingChina

Specular highlight detection and removal is a fundamental problem in computer vision and image *** this paper,we present an efficient endto-end deep learning model for automatically detecting and removing specular highlights in a single *** particular,an encoder–decoder network is utilized to detect specular highlights,and then a novel Unet-Transformer network performs highlight removal;we append transformer modules instead of feature maps in the Unet *** also introduce a highlight detection module as a mask to guide the removal ***,these two networks can be jointly trained in an effective *** to the hierarchical and global properties of the transformer mechanism,our framework is able to establish relationships between continuous self-attention layers,making it possible to directly model the mapping between the diffuse area and the specular highlight area,and reduce indeterminacy within areas containing strong specular highlight *** on public benchmark and real-world images demonstrate that our approach outperforms state-of-the-art methods for both highlight detection and removal tasks.

关键词： specular highlight detection specular highlight removal Unet-Transformer

Scene text removal via cascaded text stroke detection and erasing

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computational Visual Media 2022年第2期8卷 273-287页

作者： Xuewei Bian Chaoqun Wang Weize Quan Juntao Ye Xiaopeng Zhang dong-ming yan National Laboratory of Pattern Recognition Institute of AutomationBeijing 100049China School of Artificial Intelligence University of Chinese Academy of SciencesBeijing 100084China

Recent learning-based approaches show promising performance improvement for the scene text removal task but usually leave several remnants of text and provide visually unpleasant *** this work,a novel end-to-end framework is proposed based on accurate text stroke ***,the text removal problem is decoupled into text stroke detection and stroke removal;we design separate networks to solve these two subproblems,the latter being a generative *** two networks are combined as a processing unit,which is cascaded to obtain our final model for text *** results demonstrate that the proposed method substantially outperforms the state-of-the-art for locating and erasing scene text.A new large-scale real-world dataset with 12,120 images has been constructed and is being made available to facilitate research,as current publicly available datasets are mainly synthetic so cannot properly measure the performance of different methods.

关键词： scene text removal text stroke detection generative adversarial networks cascaded network design real-world dataset

Surface remeshing with robust user-guided segmentation

在线全文

学校读者我要写书评

暂无评论

Computational Visual Media 2018年第2期4卷 113-122页

作者： Dawar Khan dong-ming yan Fan Ding Yixin Zhuang Xiaopeng Zhang National Laboratory of Pattern Recognition Institute of AutomationChinese Academy of SciencesBeijing 100190China University of Chinese Academy of Sciences Beijing100049China National Digital Switching System Engineering&Technological Research Center China

Surface remeshing is widely required in modeling, animation, simulation, and many other computer graphics applications. Improving the elements' quality is a challenging task in surface remeshing. Existing methods often fail to efficiently remove poor-quality elements especially in regions with sharp features. In this paper, we propose and use a robust segmentation method followed by remeshing the segmented mesh. Mesh segmentation is initiated using an existing Live-wire interaction approach and is further refined using local mesh operations. The refined segmented mesh is finally sent to the remeshing pipeline, in which each mesh segment is remeshed independently. An experimental study compares our mesh segmentation method as well as remeshing results with representative existing methods. We demonstrate that the proposed segmentation method is robust and suitable for remeshing.

关键词： mesh generation surface remeshing mesh segmentation triangulation

An anisotropic Chebyshev descriptor and its optimization for deformable shape correspondence

在线全文

学校读者我要写书评

暂无评论

Computational Visual Media 2023年第3期9卷 461-477页

作者： Shengjun Liu Hongyan Liu Wang Chen dong-ming yan Ling Hu Xinru Liu Qinsong Li School of Mathematics and Statistics Central South UniversityChangsha 410000China State Key Laboratory of Multimodal Artificial Intelligence Systems Institute of AutomationChinese Academy of SciencesBeijing 100190China School of Artificial Intelligence University of Chinese Academy of SciencesBeijing 101408China School of Mathematics and Statistics Hunan First Normal UniversityChangsha 410000China Big Data Institute Central South UniversityChangsha 410000China

Shape descriptors have recently gained popularity in shape matching,statistical shape modeling,*** discriminative ability and efficiency play a decisive role in these *** this paper,we first propose a novel handcrafted anisotropic spectral descriptor using Chebyshev polynomials,called the anisotropic Chebyshev descriptor(ACD);it can effectively capture shape features in multiple *** ACD inherits many good characteristics of spectral descriptors,such as being intrinsic,robust to changes in surface discretization,***,due to the orthogonality of Chebyshev polynomials,the ACD is compact and can disambiguate intrinsic symmetry sinces everal directions are *** improve the ACD’s discrimination ability,we construct a Chebyshev spectral manifold convolutional neural network(CSMCNN)that optimizes the ACD and produces a learned *** experimental results show that the ACD outperforms existing state-of-the-art handcrafted *** combination of the ACD and the CSMCNN is better than other state-of-the-art learned descriptors in terms of discrimination,efficiency,and robustness to changes in shape resolution and discretization.

关键词： anisotropic descriptor spectral descriptor shape descriptor shape matching spectral convolution deep learning

Analyzing surface sampling patterns using the localized pair correlation function

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computational Visual Media 2016年第3期2卷 219-230页

作者： Weize Quan Jianwei Guo dong-ming yan Weiliang Meng Xiaopeng Zhang NLPR-LIAMA Institute of Automation Chinese Academy of Sciences

Point distributions with different characteristics have a crucial influence on graphics applications. Various analysis tools have been developed in recent years, mainly for blue noise sampling in Euclidean domains. In this paper, we present a new method to analyze the properties of general sampling patterns that are distributed on mesh surfaces. The core idea is to generalize to surfaces the pair correlation function(PCF) which has successfully been employed in sampling pattern analysis and synthesis in 2D and 3D. Experimental results demonstrate that the proposed approach can reveal correlations of point sets generated by a wide range of sampling algorithms. An acceleration technique is also suggested to improve the performance of the PCF.

关键词： point distribution spectral analysis pair correlation function(PCF) mesh surface

Odd-even Effect on Liquid Crystalline Behavior for C_3-Cyclotriveratrylene(CTV) Derivatives Containing Alkoxies with Different Lengths as Peripheral Groups

在线全文

学校读者我要写书评

暂无评论

Chemical Research in Chinese Universities 2009年第4期25卷 579-584页

作者： CHEN Dan-mei dong yan-ming SHEN Bing-xing yanG Xue-hui LI yan-jie HU Xiao-lan Department of Materials Science and Engineering Xiamen University Xiamen 361005 P. R. China College of Chemistry and Chemical Engineering Xiamen University Xiamen 361005 P. R. China

A series of bowlic cyclotriveratrylenes（CTV） with peripheral groups with different lengths were *** compounds were investigated by diferential scanning calorimetry and hot stage coupled polarizing *** CTV derivatives show thermotropic liquid crystalline *** experimental results of their thermotropic liquid crystalline behavior indicate that the clear points,the entropy changes of melting points,the crystallization temperatures,and their entropy changes all exhibited an evident odd-even effect except the melting points,which decreased monotonously with the increase of the length of the alkoxy *** parameter values of CTVs with even number carbon atoms were larger than those of CTVs with odd *** the length of alkoxyl chains was even longer,a monotonous decrease ***,in the case of the entropy changes of both melting points and crystallization temperatures,the effect was valid for all the six species,and therefore,the whole curves presented as a zig-zag form.

关键词： Cyclotriveratryleues(CTV) Bowlic liquid crystal Thermotropic liquid crystal Odd-even effect