检索结果-南通市图书馆

End-to-end spatial transform face detection and recognition

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Virtual Reality & Intelligent Hardware 2020年第2期2卷 119-131页

作者： Hongxin ZHANG Liying CHI cad&CG State Key lab Zhejiang UniversityHangzhou 310058China

Background Several face detection and recogni tion methods have been proposed in the past decades that have excellent *** conventional face recognition pipeline comprises the following:(1)face detection,(2)face alignment,(3)feature extraction,and(4)similarity,which are independent of each *** separate facial analysis stages lead to redundant model calculations,and are difficult for use in end-to-end *** In this paper,we propose a novel end-to-end trainable convolutional network framework for face detection and recognition,in which a geometric transformation matrix is directly learned to align the faces rather than predicting the facial *** the training stage,our single CNN model is supervised only by face bounding boxes and personal identities,which are publicly available from WIDER FACE and CASIA-WebFace *** model is tested on Face Detection Dataset and Benchmark(FDDB)and labeled Face in the Wild(LFW)*** The results show 89.24%recall for face detection tasks and 98.63%accura cy for face recognition tasks.

关键词： Face detection Face recognition Spatial transform Feature fusion

Quantum search with prior knowledge

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年第9期67卷 321-329页

作者： Xiaoyu HE Xiaoming SUN Jialing ZHANG State Key lab of Processors Institute of Computing Technology Chinese Academy of Sciences School of Computer Science and Technology University of Chinese Academy of Sciences

The combination of contextual side information and search is a powerful paradigm in the scope of artificial intelligence. The prior knowledge enables the identification of possible solutions but may be imperfect. Contextual information can arise naturally, for example in game AI where prior knowledge is used to bias move decisions. In this work we investigate the problem of taking quantum advantage of contextual information, especially searching with prior knowledge. We propose a new generalization of Grover's search algorithm that achieves the optimal expected success probability of finding the solution if the number of queries is fixed. Experiments on small-scale quantum circuits verify the advantage of our algorithm. Since contextual information exists widely, our method has wide applications. We take game tree search as an example.

关键词： quantum computing quantum search quantum query algorithm prior knowledge

EasySVM: A visual analysis approach for open-box support vector machines

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Computational Visual Media 2017年第2期3卷 161-175页

作者： Yuxin Ma Wei Chen Xiaohong Ma Jiayi Xu Xinxin Huang Ross Maciejewski Anthony K.H.Tung State Key lab of cad&CG Zhejiang University Arizona State University National University of Singapore

Support vector machines(SVMs) are supervised learning models traditionally employed for classification and regression analysis. In classification analysis, a set of training data is chosen, and each instance in the training data is assigned a categorical class. An SVM then constructs a model based on a separating plane that maximizes the margin between different classes. Despite being one of the most popular classification models because of its strong performance empirically, understanding the knowledge captured in an SVM remains difficult. SVMs are typically applied in a black-box manner where the details of parameter tuning, training, and even the final constructed model are hidden from the users. This is natural since these details are often complex and difficult to understand without proper visualization tools. However, such an approach often brings about various problems including trial-and-error tuning and suspicious users who are forced to trust these models *** contribution of this paper is a visual analysis approach for building SVMs in an open-box *** goal is to improve an analyst's understanding of the SVM modeling process through a suite of visualization techniques that allow users to have full interactive visual control over the entire SVM training *** visual exploration tools have been developed to enable intuitive parameter tuning, training datamanipulation, and rule extraction as part of the SVM training process. To demonstrate the efficacy of our approach, we conduct a case study using a real-world robot control dataset.

关键词： support vector machines(SVMs) rule extraction visual classification high-dimensional visualization visual analysis

在线全文

学校读者我要写书评

暂无评论

Path Sensitization

Journal of Computer Science & Technology 1997年第3期12卷 271-282页

作者：赵行闵应骅李忠诚 Center for Fault-Tolerant Computing CAD Lab. Institute of Computing Technology Chinese Academy of Sciences Beijing

For different delay models, the concept of sensitization can be very *** concepts of sensitization cannot precisely describe circult behavior when the input vectors change very fast. Using Boolean process approach, this paper presents a new definition of sensitization for arbitrary input *** this new concept it is found that if the inputs of a combinational circult can change at any time, and each gate's delay varies witliln an interval (bounded gate delay model), then every path, which is not necessarily a single topological path, is sensitizable. fum the experimental results it can be seen that, all nonsensitizable paths for traditional concepts actually can propagate transitions along them for some input wavforms. However, specilied time between input transitions (STBIT) and nilnimum perndssible pulse width (E) are two major factors to make some paths non-sensitizable.

关键词： Sensitization delay model Boolean process waveform

A Robust and Fast Non-Local Means Algorithm for Image Denoising

在线全文

学校读者我要写书评

暂无评论

Journal of Computer Science & Technology 2008年第2期23卷 270-279页

作者：刘艳丽王进陈曦郭延文彭群生 State Key lab of cad & CG Zhejiang UniversityHangzhou 310058China State Key lab of Novel Software Technology Nanjing UniversityNanjing 210000China

In the paper, we propose a robust and fast image denoising method. The approach integrates both Non- Local means algorithm and Laplacian Pyramid. Given an image to be denoised, we first decompose it into Laplacian pyramid. Exploiting the redundancy property of Laplacian pyramid, we then perform non-local means on every level image of Laplacian pyramid. Essentially, we use the similarity of image features in Laplacian pyramid to act as weight to denoise image. Since the features extracted in Laplacian pyramid are localized in spatial position and scale, they are much more able to describe image, and computing the similarity between them is more reasonable and more robust. Also, based on the efficient Summed Square Image （SSI） scheme and Fast Fourier Transform （FFT）, we present an accelerating algorithm to break the bottleneck of non-local means algorithm - similarity computation of compare windows. After speedup, our algorithm is fifty times faster than original non-local means algorithm. Experiments demonstrated the effectiveness of our algorithm.

关键词： image denoising non-local means Laplacian pyramid summed square image FFT

***:A web-based library for high-efficiency visualization of large-scale graphs and networks

在线全文

学校读者我要写书评

暂无评论

Visual Informatics 2021年第1期5卷 61-66页

作者： Dongming Han Jiacheng Pan Xiaodong Zhao Wei Chen State Key lab of cad&CG Zhejiang UniversityHangzhouZhejiangChina Zhejiang lab HangzhouZhejiangChina

Graph visualization plays an important role in several fields,such as social media networks,protein-protein interaction networks,and traffic networks.A number of visualization design tools and programming toolkits have been widely used in graph-related ***,a key challenge remains in the high-efficiency visualization of large-scale graph *** this study,we present ***,an open-source and WebGL-based JavaScript library that supports the fast visualization of large-scale graph data(up to 50 thousand nodes and 1 million edges)at an interactive frame rate with a commodity *** results demonstrate that our library outperforms existing toolkits(***,***,***,and ***)in terms of performance.

关键词： Graph Graph visualization Network visualization Node-link diagram

Volumetric data modeling and analysis based on seven-directional box spline

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2014年第6期57卷 260-273页

作者： FANG MeiE LU Jia PENG QunSheng School of Computer Science and technology Hangzhou Dianzi University State Key lab of cad&CG Zhejiang University

The existing methods for visualizing volumetric data are mostly based on piecewise linear *** all kinds of analysis based on them have to be substituted by coarse *** both accuracy and reliability of the traditional framework for visualization and analysis of volumetric data are far from our needs of digging information implied in volumetric data *** this paper,we propose a novel framework based on a C2-continuous seven-directional box spline,under which reconstruction is of high accuracy and differential computations relative to analysis based on the reconstruction model are *** introduce a polynomial differential operator to improve the reconstruction *** order to settle the difficulty of evaluating upon the seven-directional box spline,we convert it into B′ezier form and propose effective theories and algorithms of extracting iso-surfaces,critical points and *** of examples are also given in this paper to illustrate that the novel framework is suitable for analysis,the improved reconstruction method has high accuracy,and our algorithms are fast and stable.

关键词： volumetric data visualization and analysis seven-directional box spline iso-surface critical points curvature

Direction monotonicity for a rational Bézier curve

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Applied Mathematics(A Journal of Chinese Universities) 2016年第1期31卷 1-20页

作者： SHEN Wan-qiang WANG Guo-zhao HUANG Fang School of Science Jiangnan University State Key lab of cad & CG Zhejiang University Department of Mathematics Zhejiang University

The monotonicity of a rational Bezier curve, usually related to an explicit function, is determined by the used coordinate system. However, the shape of the curve is independent of the coordinate system. To meet the affine invariant property, a kind of generalized mono- tonicity, called direction monotonicity, is introduced for rational Bezier curves. The direction monotonicity is applied to both planar and space curves and to both Cartesian and affine co- ordinate systems, and it includes the traditional monotonicity as a subcase. By means of it, proper affine coordinate systems may be chosen to make some rational Bezier curves monotonic. Direction monotonic interpolation may be realized for some of the traditionally nonmonotonic data as well.

关键词： rational Bezier curve monotonicity explicit function affine coordinate system interpolation.

An improved parallel contrast-aware halftoning

在线全文

学校读者我要写书评

暂无评论

Journal of Zhejiang University-Science C(Computers and Electronics) 2013年第12期14卷 918-929页

作者： Ling-yue LIU Wei CHEN Tien-tsin WONG Wen-ting ZHENG Wei-dong GENG State Key lab of cad & CG Zhejiang University Department of Computer Science and Engineering The Chinese University of Hong Kong

Digital image halftoning is a widely used technique. However, achieving high fidelity tone reproduction and structural preservation with low computational time cost remains a challenging problem. This paper presents a highly parallel algorithm to boost real-time application of serial structure-preserving error diffusion. The contrast-aware halftoning approach is one such technique with superior structure preservation, but it offers only a limited opportunity for graphics processing unit(GPU) acceleration. Our method integrates contrast-aware halftoning into a new parallelizable error-diffusion halftoning framework. To eliminate visually disturbing artifacts resulting from parallelization, we propose a novel multiple quantization model and space-filling curve to maintain tone consistency, blue-noise property, and structure consistency. Our GPU implementation on a commodity personal computer achieves a real-time performance for a moderately sized image. We demonstrate the high quality and performance of the proposed approach with a variety of examples, and provide comparisons with state-of-the-art methods.

关键词： Digital image halftoning Error diffusion GPU Quantization Space-filling curve