|
|
|
3D model semantic classification and retrieval with
Gaussian processes |
| GAO Bo-yong1,2, YAO Fu-tian1,2, ZHANG San-yuan1 |
1. Department of Computer Science and Engineering, Zhejiang University, Hangzhou 310027, China;
2. Department of Computer Science and Technology, China Jiliang University, Hangzhou 310018, China |
|
|
|
Abstract A novel 3D model retrieval and semantic classification method using Gaussian processes was proposed to improve the performance of 3D model retrieval systems, and reduce the ‘semantic gap’ between the shape features and the richness of human semantics. A new type of feature named AC2 using histogram of angle between the centroid and pairs of random points was proposed, which combined D2 of shape distribute as low-level feature. The Gaussian processes were used for 3D model semantic classification as supervised learning, and the predictive distribution of the semantic class probability was computed for associating low-level features with query concepts. The method ranked models by dissimilarity measure incorporating the semantic distance and the shape feature distance. Experimental results showed that the multi-class 3D model classification accuracy using the proposed method is significantly higher than those of other supervised learning methods, and the retrieval can capture the query model’s semantics, so the performance is improved.
|
|
Published: 01 December 2010
|
|
|
|
Corresponding Authors:
张三元,男,教授,博导.
E-mail: syzhang@zju.edu.cn
|
基于高斯过程的三维模型语义分类和检索
针对三维模型检索系统提高准确率、减少几何特征和人类语义丰富性之间的“语义鸿沟”等问题, 提出一种基于高斯过程的语义分类和检索新方法.该方法采用一种统计2个采样点相对质心向量夹角的AC2直方图新特征,与形状分布的D2特征组合成低层特征，使用高斯过程进行三维模型语义分类的监督学习,计算测试模型的语义类概率预测分布,建立低层特征和查询概念之间的联系；使用语义距离和不相似度计算方法进行检索排序.实验结果表明：与已有的某些监督学习的方法相比,多类的测试模型进行语义分类的准确率明显得到提升,检索中能体现语义概念,检索性能也得到提高.
|
|
| ［1］ VELTKAMP R C, RUIJSENAARS R, SPAGNUOLO M, et al. SHREC2006: 3D shape retrieval contest ［R］. Utrecht: Department of Information and Computing Sciences, Utrecht University, 2006.
［2］ TANGELDER J, VELTKAMP R. A survey of content based 3D shape retrieval methods ［J］. Multimedia Tools and Applications, 2008, 39(3): 441-471.
［3］ CORTELAZZO G M, ORIO N. Retrieval of colored 3D models ［C］∥ Proceedings of the Third International Symposium on 3D Data Processing, Visualization, and Transmission (3DPVT′06). Washington: IEEE Computer Society, 2006: 986-993.
［4］ HILAGA M, SHINAGAWA Y, KOHMURA T, et al. Topology matching for fully automatic similarity estimation of 3D shapes ［C］∥ Proceedings of the 28th annual conference on Computer graphics and interactive techniques. New York: ACM, 2001: 203-212.
［5］ FILALI ANSARY T, DAOUDI M, VANDEBORRE J. A Bayesian 3D search engine using adaptive views clustering ［J］. IEEE Transactions on Multimedia, 2007, 9(1): 78-88.
［6］ ELAD M, TAL A, AR S. Content based retrieval of VRML objects: an iterative and interactive approach ［C］∥ Proceedings of the Sixth Eurographics Workshop on Multimedia 2001. Vienna: SpringerVerlag GmbH, 2002: 107-118.
［7］ AKGL C B, SANKUR B, YEMEZ Y, et al. Similarity score fusion by ranking risk minimization for 3D object retrieval ［C］∥ Proceedings of the Eurographics Workshop on 3D Object Retrieval(2008). Crete: ［s. n.］, 2008: 1-9.
［8］ ATMOSUKARTO I, LEOW W K, HUANG Z. Feature combination and relevance feedback for 3D model retrieval ［C］∥ Proceedings of the 11th International Multimedia Modelling Conference. Washington: IEEE Computer Society, 2005: 334-339.
［9］ FUNKHOUSER T, SHILANE P. Partial matching of 3D shapes with prioritydriven search ［C］∥ Proceedings of the Fourth Eurographics Symposium on Geometry Processing. AirelaVille: Eurographics Association, 2006: 131-142.
［10］ OHBUCHI R, YAMAMOTO A, KOBAYASHI J. Learning semantic categories for 3D model retrieval ［C］∥ Proceedings of the International Workshop on Workshop on Multimedia Information Retrieval. New York: ACM, 2007: 31-40.
［11］ LEIFMAN G, MEIR R, TAL A. Semanticoriented 3d shape retrieval using relevance feedback ［J］. The Visual Computer, 2005, 21(8): 865-875.
［12］ LU Ke, ZHAO Feng, HE Ning. An effective approach to contentBased 3D model retrieval and classification ［C］∥ Proceedings of the 2007 International Conference on Computational Intelligence and Security. Washington: IEEE Computer Society, 2007: 361-365.
［13］ 陈磊,潘翔,叶修梓,等.基于本体的产品知识表达和检索技术研究［J］.浙江大学学报：工学版,2008,42(12): 2037-2042, 2048.
CHEN Lei, PAN Xiang, YE Xiuzi, et al. Ontologybased product knowledge representation and retrieval ［J］. Journal of Zhejiang University ： Engineering Science, 2008, 42(12): 2037-2042, 2048.
［14］ XU Dong, LI Hua. 3D shape retrieval integrated with classification information ［C］∥ Proceedings of the Fourth International Conference on Image and Graphics. Washington: IEEE Computer Society, 2007: 774-779.
［15］ VAPNIK V N. The nature of statistical learning theory ［M］. 2nd ed. New York: Springer, 2000.
［16］ OSADA R, FUNKHOUSER T, CHAZELLE B, et al. Shape distributions ［J］. ACM Transactions on Graphics, 2002, 21 (4): 807-832.
［17］ HUGHES G. On the mean accuracy of statistical pattern recognizers ［J］. IEEE Transactions on Information Theory, 1968, 14(1): 55-63.
［18］ RASMUSSEN C E, WILLIAMS C K I. Gaussian processes for machine learning ［M］. Cambridge/London: The MIT Press, 2006.
［19］ SHILANE P, MIN P, KAZHDAN M, et al. The Princeton shape benchmark ［C］∥ Proceedings of the Shape Modeling International 2004. Washington: IEEE Computer Society, 2004: 167-178.
［20］ 刘小明,尹建伟,冯志林,等.基于适应加权非对称AdaBoost HMM的三维模型分类算法［J］.浙江大学学报：工学版,2006,40(8): 1300-1305.
LIU Xiaoming, YIN Jianwei, FENG Zhilin, et al. 3D model classification based on adaptiveweighted asymmetric AdaBoost hidden Markov models ［J］. Journal of Zhejiang University ：Engineering Science, 2006, 40(8): 1300-1305. |
|
Viewed |
|
|
|
Full text
|
|
|
|
|
Abstract
|
|
|
|
|
Cited |
|
|
|
|
| |
Shared |
|
|
|
|
| |
Discussed |
|
|
|
|
