logo

字形计算实验室

Zhouhui Lian (连宙辉)










I am currently an Associate Professor at the Institute of Computer Science and Technology, Peking University, Beijing, China.

Character Shape Computing Lab: http://59.108.48.27/

Try to create your own handwriting fonts: http://59.108.48.27/flexifont  (For free!)

Recent Events:


—— 2017 ——— 

2017-03: Becomes the PC members of three conferences (ICIP, ICDAR, EG 3DOR)

2017-02-28: One paper is accepted by Eurographics 2017 as the oral short paper

2017-02-27: Two Papers are accepted by CVPR 2017 as full papers


—— 2015 ——— 

Contest: "SHREC'15 —- Non-rigid 3D Shape Retrieval"

Organizer: Zhouhui Lian, Jun Zhang

Password to access the database: SHREC15@PKU

—— 2012 ——— 

2012-08-20: Get a research funding from National Natural Science Foundation of China (NSFC)

2012-07-20: Our paper "A comparison of methods for non-rigid 3D shape retrieval" is accepted by Pattern Recognition

2012-07-16: Our paper "Feature-preserved 3D Canonical Form" is accepted by International Journal of Computer Vision (IJCV)

2012-03-11: Our paper " A New Convexity Measurement for 3D Meshes" is accepted by CVPR 2012

—— 2011 ——–

Contest: "SHREC'11 —- Non-rigid 3D Watertight Shape Retrieval"

Organizer: Zhouhui Lian, Afzal Godil 

Password to access the database: SHREC11@NIST

Toolkit of the benchmark

Contact Details

Address: No 128 ZhongGuanCun North Street, HaiDian District, Beijing 100080, P.R.China

Email: lianzhouhui (at) pku.edu.cn

Education

09/2001 ~ 07/2005                          College of Automation Engineering, Nanjing University of Aeronautics and Astronautics (NUAA), China          Bachelor Degree

09/2005 ~ 06/2011                          School of Automation Science and Electrical Engineering, Beihang University (BUAA), China                          PhD Degree

Professional Experience

01/2008 ~ 01/2009                          School of Computer Science, Cardiff University, UK                                                                                           Academic Visitor

07/2009 ~ 07/2011                          Information Access Division, National Institute of Standards and Technology, USA                                          Guest Researcher

07/2011 ~ Now                                Institute of Computer Science and Technology, Peking University, China                                                          PostDoc Researcher 

Research Interests

Computer Vision

Computer Graphics

Remote Sensing

Control Engineering

Projects

National Natural Science Fondation of China: "Researches on Automatic Generation of High-quality Chinese Fonts based on Shape Analysis and Processing" (Project Leader)

China Postdoctoral Science Foundation Project: "Researches on Several Key Technologies of Non-rigid 3D Shape Retrieval" (Project Leader)

Doctoral Graduate Research Program: "3D Shape Retrieval"

NIST Research Project: "SHARP – Shape Analysis Research Project"

Shape Retrieval Contest 2010: "SHREC'10 —- Non-rigid 3D Shape Retrieval" (Organizer) 

Shape Retrieval Contest 2011: "SHREC'11 —- Non-rigid 3D Watertight Shape Retrieval" (Organizer) 

China National 863 Plan Project: “Research of the key technology of a novel Aerial Camera"

 

Selected Publications (Including 2 IJCV, 1 CVPR, and 1 Pattern Recognition papers)

 

 

Feature-preserved 3D Canonical Form

International Journal of Computer Vision (IJCV), 2013, vol. 102, no. 1-3, pp. 221-238

Zhouhui Lian, Afzal Godil, Jianguo Xiao

(Data: PKUNSB database, Distance Matrices of 5 Methods)

 

Abstract: Measuring the dissimilarity between non-rigid objects is a challenging problem in 3D shape retrieval. One potential solution is to construct the models’ 3D canonical forms (i.e., isometry-invariant representations in 3D Euclidean space) on which any rigid shape matching algorithm can be applied. However, existing methods, which are typically based on embedding procedures, result in greatly distorted canonical forms, and thus could not provide satisfactory performance to distinguish non-rigid models.

 

In this paper, we present a feature-preserved canonical form for non-rigid 3D meshes. The basic idea is to naturally deform original models against corresponding initial canonical forms calculated by Multidimensional Scaling (MDS). Specifically, objects are first segmented into nearrigid subparts, and then, through properly-designed  rotations and translations, original subparts are transformed into poses that correspond well with their positions and directions on MDS canonical forms. Final results are obtained by solving some nonlinear minimization problems for optimal alignments and smoothing boundaries between subparts. Experiments on a widely utilized non-rigid 3D shape benchmark not only verify the advantages of our algorithm against existing approaches, but also demonstrate that, with the help of the proposed canonical form, we can obtain significantly better retrieval accuracy compared to the state of the art.

 

A comparison of methods for non-rigid 3D shape retrieval

Pattern Recognition, Vol 46, Num 1, Pages 449-461, 2013

Zhouhui Lian, Afzal Godil, et al.

(Data: Dataset, Toolkit, Distance matrices of the methods compared)


Abstract: Non-rigid 3D shape retrieval has become an active and important research topic in content-based 3D object retrieval. The aim of this paper is to measure and compare the performance of state-of-the-art methods for non-rigid 3D shape retrieval. The paper develops a new benchmark consisting of 600 non-rigid 3D watertight meshes, which are equally classified into 30 categories, to carry out experiments for 11 different algorithms, whose retrieval accuracies are evaluated using 6 commonly-utilized measures. Models and evaluation tools of the new benchmark are publicly available on our web site.

  

Automatic Shape Morphing for Chinese Characters 

SIGGRAPH Asia 2012 (technical briefs),  2012

Zhouhui Lian, Jianguo Xiao


Abstract: How to automatically implement shape morphing for Chinese characters represented in different styles is a challenging task. In this paper, we propose a novel method to solve this problem.

Specifically, we first generate the shape template, which includes a skeleton, strokes, key points, and connection triangles, for every character in a standard Chinese font (i.e., Kaiti) library. Then, we decompose two given Chinese characters into strokes to establish an accurate correspondence between them by applying the Coherent Point Drift (CPD) algorithm to achieve non-rigid point set

registration between each character and the corresponding template. Finally, we construct an isomorphic triangulation for the source and target character shapes, and then apply as-rigid-as-possible shape interpolation on these two triangle meshes. Experimental results demonstrate the effectiveness of our method. 

 

A New Convexity Measurement for 3D Meshes

IEEE International Conference on Computer Vision and Pattern Recognition (CVPR), 119-126, 2012

Zhouhui Lian, Afzal Godil, Paul Rosin, Xianfang Sun

 

Abstract: This paper presents a novel convexity measurement for 3D meshes. The new convexity measure is calculated by minimizing the ratio of the summed area of valid regions in a mesh's six views, which are projected on faces of the bounding box whose edges are parallel to the coordinate axes, to the sum of three orthogonal projected areas of the mesh. The complete definition, theoretical analysis, and a computing algorithm of our convexity measure are explicitly described. This paper also proposes a new 3D shape descriptor CD (i.e., Convexity Distribution) based on the distribution of above-mentioned ratios, which are computed by randomly rotating the mesh around its center, to better describe the object's convexity-related properties compared to existing convexity measurements. Our experiments not only show that the proposed convexity measure corresponds well with human intuition, but also demonstrate the effectiveness of the new convexity measure and the new shape descriptor by significantly improving the performance of other methods in the application of 3D shape retrieval.

  

 

 

Visual Similarity based 3D Shape Retrieval Using Bag-of-Features 

Shape Modeling International 2010 (SMI'10), 25-36, 2010, (Oral Presentation)

Zhouhui Lian, Afzal Godil, Xianfang Sun

 

Abstract: This paper presents a novel 3D shape retrieval method, which uses Bag-of-Features and an efficient multi-view shape matching scheme. In our approach, a properly normalized object is first described by a set of depth-buffer views captured on the surrounding vertices of a given unit geodesic sphere. We then represent each view as a word histogram generated by the vector quantization of the view’s salient local features. The dissimilarity between two 3D models is measured by the minimum distance of their all (24) possible matching pairs.

This paper also investigates several critical issues including the influence of the number of views, codebook, training data, and distance function. Experiments on four commonly-used benchmarks demonstrate that: 1) Our approach obtains superior performance in searching for rigid models. 2) The local feature and global feature based methods are somehow complementary. Moreover, a linear combination of them significantly outperforms the state-of-the-art in terms of retrieval accuracy.

 

Full List (* means corresponding author)

 

Selected Journal Papers：

[1] D. Pickup, X. Sun, P. L. Rosin, R. R. Martin, Z. Cheng, Z. Lian, M. Aono, A. Ben Hamza, A. Bronstein, M. Bronstein, S. Bu, U. Castellani, S. Cheng, V. Garro, A. Giachetti, A. Godil, L. Isaia, J. Han, H. Johan, L. Lai, B. Li, C. Li, H. Li, R. Litman, X. Liu, Z. Liu, Y. Lu, L. Sun, G. Tam, A. Tatsuma, J. Yeshow less. Shape Retrieval of Non-rigid 3D Human Models. International Journal of Computer Vision (IJCV). 2016.

[2] Z. Lian*, A. Godil, J. Xiao. Feature-preserved 3D Canonical Form, International Journal of computer Vision (IJCV), vol. 102, no. 1-3, pp. 221-238, 2013

[3] Z. Lian*, A. Godil, B. Bustos, M. Daoudi, J. Hermans, S. Kawamura, Y. Kurita, G. Lavoué, H.V. Nguyen, R. Ohbuchi, Y. Ohkita, Y. Ohishi, F. Porikli, M. Reuter, I. Sipiran, D. Smeets, P. Suetens, H. Tabia, D. Vandermeulen. A Comparison of Methods for Non-rigid 3D Shape Retrieval, Pattern Recognition (PR), vol. 46, no. 1, pp.449-461, 2013

[4] Z. Lian*, A. Godil, X. Sun, J. Xiao. CM-BOF: Visual Similarity based 3D Shape Retrieval Using Clock Matching and Bag-of-Features, Machine Vision and Applications (MVAP), vol. 24, no. 8, pp. 1685-1704, 2013

[5] Z. Lian*, P.L. Rosin, X. Sun. Rectilinearity of 3D meshes, International Journal of Computer Vision (IJCV), vol. 89, no. 2-3, pp. 130-151, 2010


Selected Conference Papers：

[1] J. Liu, Z. Lian*, Y. Wang, J. Xiao. Incremental Kernel Null Space Discriminant Analysis for Novelty Detection. CVPR 2017 (accepted) (Data & Source code)

[2] S. Yang, J. Liu, Z. Lian, Z. Guo. Awesome Typography: Statistics-Based Text Effects Transfer. CVPR 2017 (accepted)

[3] Z. Lian*, B. Zhao, J. Xiao. Automatic Generation of Large-scale Handwriting Fonts via Style Learning. Siggraph Asia 2016, Artical no. 12 (Technical briefs), 2016

[4] R. Sun, Z. Lian*, Y. Tang, J. Xiao. Aesthetic Visual Quality Evaluation of Chinese Handwritings. IJCAI 2015, pp. 2510-2516, 2015 (Data & Source code)

[5] J. Liu, Z. Lian, J. Feng, B. Zhou. Sketch based Modeling via Manifold Regularization. Siggraph Asia 2015, Article no. 15 (Technical briefs), 2015 (Data & Source code)

[6] Z. Lian*, J. Xiao. Automatic Shape Morphing for Chinese Characters, Siggraph Asia 2012, Article no. 2 (Technical briefs), 2012

[7] Z. Lian*, A. Godil, P.L. Rosin, X. Sun. A New Convexity Measurement for 3D Meshes, CVPR 2012, pp. 119-126, 2012


Other Papers：

[1] X. Chen, Z. Lian*, Y. Tang, J. Xiao. An Automatic Stroke Extraction Method using Manifold Learning. Eurographics 2017 (oral, short paper) (accepted)

[2] D. Liu, Z. Lian*, Y. Tang, J. Xiao. Structure-aware Image Resizing for Chinese Characters. MMM 2017 (oral, full paper) (accepted)

[3] W. Dong, Z. Lian*, Y. Tang, J. Xiao. Text Detection in Natural Images Using Localized Stroke Width Transform. MMM 2015, pp. 49-58, 2015

[4] J. Zhang, Z. Lian*, Z. Liu, J. Xiao. CEFM: A Heuristic Mesh Segmentation Method based on Convexity Estimation and Fast Marching. International Conference on Computer Graphics Theory and Applications (GRAPP 2015), pp. 114-121, 2015

[5] Z. Lian*, J. Zhang, et al. SHREC’15 Track: Non-rigid 3D Shape Retrieval. Eurographics Workshop on 3D Object Retrieval (3DOR), pp. 107-120, 2015

[6] W. Song, Z. Lian*, Y. Tang, J. Xiao. Content-Independent Font Recognition on a Single Chinese Character using Sparse Representation. ICDAR 2015, pp. 376-380, 2015

[7] X. Chen, Z. Lian*, Y. Tang, J. Xiao. A Benchmark For Stroke Extraction of Chinese Characters. Acta Scientiarum Naturalium Universitatis Pekinensis, 2015

[8] Y. Yi, Z. Lian*, Y. Tang, J. Xiao. A data-driven personalized digital ink for Chinese characters. MMM 2014, pp. 254-265, 2014 

[9] W. Pan, Z. Lian*, R. Sun, Y. Tang, J. Xiao. FlexiFont: A Flexible System to Generate Personal Font Libraries. DocEng 2014, pp. 17-20, 2014 

[10] W. Pan, Z. Lian*, Y. Tang, J. Xiao. Skeleton-Guided Vectorization of Chinese Calligraphy Images. MMSP 2014, paperID 19, 2014 

[11] H. Sun, Z. Lian*, Y. Tang, J. Xiao. Non-Rigid Point Set Registration For Chinese Characters Using Structure-Guided Coherent Point Drift. ICIP 2014, pp. 4752-4756, 2014 

[12] 易天旸,连宙辉*,孙浩,唐英敏,肖建国. 一种新的汉字笔画自动提取算法及其应用. 第十届中国计算机图形学大会(ChinaGraph 2014).

[13] C. Wang, Z. Lian*, Y. Tang, J. Xiao. Automatic Correspondence Finding for Chinese Characters using Graph Matching. The Seventh International Conference on Image and Graphics (ICIG 2013), pp. 545-550, 2013  

[14] 孙浩,唐英敏,连宙辉,肖建国. 汉字部件的无失真缩放变换方法研究. 计算机应用研究. 2012年10月

[15] Z. Lian*, A. Godil. A Feature-preserved Canonical Form for 3D Meshes, International Conference on 3D Imaging, Modeling, Processing, Visualization and Transmission (3DIMPVT), pp. 116-123. 2011(Oral)

[16] Z. Lian*, A. Godil, B. Bustos, M. Daoudi, J. Hermans, S. Kawamura, Y. Kurita, G. Lavoué, H.V. Nguyen, R. Ohbuchi, Y. Ohkita, Y. Ohishi, F. Porikli, M. Reuter, I. Sipiran, D. Smeets, P. Suetens, H. Tabia, D. Vandermeulen. SHREC’11 Track: Shape Retrieval on Non-rigid 3D Watertight Meshes. Eurographics Workshop on 3D Object Retrieval (3DOR), pp. 79-88, 2011

[17] Z. Lian*, A. Godil, X. Sun. Visual Similarity Based 3D Shape Retrieval Using Bag-of-Features, Shape Modeling International (SMI), pp. 25-36, 2010 (Oral)

[18] Z. Lian*, A. Godil, X. Sun, H. Zhang. Non-rigid 3D Shape Retrieval Using Multidimensional Scaling and Bag-of-Features, International Conference on Image Processing (ICIP), pp. 3181-3184, 2010

[19] Z. Lian*, A. Godil, T. Fabry, T. Furuya, J. Hermans, R. Ohbuchi, C. Shu, D. Smeets, P. Suetens, D. Vandermeulen, S. Wuhrer. SHREC’10 Track: Non-rigid 3D Shape Retrieval. Eurographics Workshop on 3D Object Retrieval (3DOR), pp. 101-108, 2010

[20] V.T. Porethi, A. Godil, H. Dutagaci, T. Furuya, Z. Lian, R. Ohbuchi. SHREC’10 Track: Generic 3D Warehouse. Eurographics Workshop on 3D Object Retrieval (3DOR), pp. 93-100, 2010

[21] A.M. Bronstein, M.M. Bronstein, U. Castellani, B. Falcidieno, A. Fusiello, A. Godil, L.J. Guibas, I. Kokkinos, Z. Lian, M. Ovsjanikov, G. Patanè, M. Spagnuolo, R. Toldo. SHREC’10 Track: Robust Shape Retrieval. Eurographics Workshop on 3D Object Retrieval (3DOR), pp. 71-78, 2010

[22] R.C. Veltkamp, G. Giezeman, H. Bast, T. Baumbach, T. Furuya, J. Giesen, A. Godil, Z. Lian, R. Ohbuchi, W. Saleem. SHREC’10 Track: Large-scale Retrieval. Eurographics Workshop on 3D Object Retrieval (3DOR), pp. 63-69, 2010 

[23] A. Godil, H. Dutagaci, C. Akgül, A. Axenopoulos, B. Bustos, M. Chaouch, P. Daras, T. Furuya, S. Kreft, Z. Lian, T. Napoleon, A. Mademlis, R. Ohbuchi, P. Rosin, B. Sankur, T. Schreck, X. Sun, M. Tezuka, A. Verroust-Blondet, M. Walter, Y. Yemez. SHREC’09 Track: Generic Shape Retrieval. Eurographics Workshop on 3D Object Retrieval (3DOR), pp. 62-68, 2009

[24] Z. Lian*, P.L. Rosin, X. Sun. A Rectilinearity Measurement for 3D Meshes. ACM International Conference on Multimedia Information Retrieval (MIR), pp. 395-402, 2008 (Oral)

 

Funding Experience

 

01/2013 ~ 12/2015:

Project Leader.  National Natural Science Fondation of China (NSFC Grant No.: 61202230), "Researches on Automatic Generation of High-quality Chinese Fonts based on Shape Analysis and Processing", about $36K

01/2013 ~ 12/2014:

Project Leader. Special Support of China Postdoctoral Science Foundation (Grant No.: 2013T60038), “Researches on Key Technologies of the Automatic Generation of Chinese Handwriting Fonts”, about $27K

07/2011 ~ 07/2013:

Project Leader. China Postdoctoral Science Foundation (Grant No.: 2012M510274), “Researches on Several Key Technologies of Non-rigid 3D Shape Retrieval”, about $9K

05/2006 ~ 09/2006:

Team leader of an application group, resulted in a successful proposal for a National 863 Plan grant, “Research of the key technology of a novel Aerial Camera”, about $130K (Project NO: 2006AA12Z119)

 


Honors

2014           First class award of the 14th teaching skill contest for young teachers in Peking University (top 2)  

2012           Excellent postdoctoral researcher in Peking University (top 1%) 

2006           First class academic innovation award of Spatial Information Integration & Its Applications Beijing Key Laboratory, Peking University (1/30, the only one in the lab)

2005           Third class award of Jiangsu Province’s excellent undergraduate research paper (top 1%)

2005           Excellent graduation thesis, NUAA (top 2%)

2003           First class award of Jiangsu Province, China Undergraduate Mathematical Contest in Modeling

2003           First class award, “SuHe” special Scholarship, NUAA (1/140, the only one in the department)

2002           First class award, “618” special Scholarship, NUAA (1/140, the only one in the department)

 

 

Links:

 

Positions:

Institute of Science and Technology Austria (Christoph Lampert)

Toshiba Research Europe, Cambridge, UK (Cambridge Research Laboratory)

Microsoft Cambridge Group, UK (Cambridge Research Lab)

USTB, China 

 University of Coperhagen, Denmark (image group)

 

Research Groups:

 

Codes:


回到顶部