自监督非等距三维模型簇对应关系计算方法

doi:10.16451/j.cnki.issn1003-6059.202403006

摘要
图/表
参考文献
相关文章 (15)

全文: PDF (3461 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要针对现有非等距三维模型簇对应关系计算方法准确率较低且泛化能力较差的问题,文中提出采用深度内外特征对齐算法的自监督非等距三维模型簇对应关系计算方法.首先,利用DiffusionNet直接学习原始的三维模型特征,获取具有鉴别力的特征描述符.然后,使用深度内外特征对齐算法计算非等距模型对之间的对应关系,采用局部流形调和基作为模型本征信息,并结合笛卡尔坐标等外部信息,实现内外部信息对齐一致,以无监督的方式自动生成对应结果.最后,构建非等距模型簇的加权无向图,根据相似几何模型存在内在相关性的原则,设计自监督多模型匹配算法,不断增强模型图最短路径的循环一致性,获得最优的非等距三维模型簇对应关系.实验表明,文中方法对应关系测地误差较小,结果较准确,能处理模型自身对称性影响对应关系计算的问题,具有良好的泛化能力.

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	吴衍
	杨军
	张思洋

关键词 ：对应关系, 非等距模型簇, 自监督, 深度学习, 内外特征对齐

Abstract：Aiming at the problem of low accuracy and poor generalization ability in existing non-isometric 3D shape collection correspondence calculation methods, a self-supervised non-isometric 3D shape collection correspondence calculation method using deep intrinsic-extrinsic feature alignment algorithm is proposed. Firstly, discriminative feature descriptors are obtained by directly learning the original 3D shape features through DiffusionNet. Then, the deep intrinsic-extrinsic feature alignment algorithm is employed to compute correspondences between non-isometric shapes. Consistency between internal and external information is realized by utilizing local manifold harmonic bases as intrinsic information of the shapes and integrating external information such as Cartesian coordinates. Consequently, correspondence results are generated automatically in an unsupervised manner. Finally, a weighted undirected graph of non-isometric shape collections is constructed. Based on the principle of inherent correlation among similar geometric shapes, a self-supervised multi-shape matching algorithm is designed to continuously enhance the cycle-consistency of the shortest path in the shape graph, and thus optimal correspondences for non-isometric 3D shape collections are obtained. Experimental results demonstrate that the proposed method achieves small geodesic errors in correspondences with accurate results, and effectively deals with the symmetric ambiguity problem with good generalization ability.

Key words： Correspondence Non-Isometric Shape Collection Self-Supervised Deep Learning Intrinsic-Extrinsic Feature Alignment

收稿日期: 2023-11-06

ZTFLH:

TP 391.4

基金资助:国家自然科学基金项目(No.42261067)、福建省自然科学基金项目(No.2022J01972)、2021年度中央引导地方科技发展资金项目(No.2021-51)资助

通讯作者: 杨军,博士,教授,主要研究方向为计算机图形学、遥感影像分析与处理、深度学习.E-mail:yangj@mail.lzjtu.cn.

作者简介: 吴衍,博士研究生,主要研究方向为三维模型对应关系计算、深度学习.E-mail:wuyan9053@dingtalk.com. 张思洋,硕士,主要研究方向为机器学习、软件缺陷预测.E-mail:siyang.zhang1997@outlook.com.

引用本文:

吴衍, 杨军, 张思洋. 自监督非等距三维模型簇对应关系计算方法[J]. 模式识别与人工智能, 2024, 37(3): 253-266. WU Yan, YANG Jun, ZHANG Siyang. Self-Supervised Non-Isometric 3D Shape Collection Correspondence Calculation Method. Pattern Recognition and Artificial Intelligence, 2024, 37(3): 253-266.

链接本文:

http://manu46.magtech.com.cn/Jweb_prai/CN/10.16451/j.cnki.issn1003-6059.202403006 或 http://manu46.magtech.com.cn/Jweb_prai/CN/Y2024/V37/I3/253

[1] VAN KAICK O, ZHANG H, HAMARNEH G, et al. A Survey on Shape Correspondence. Computer Graphics Forum, 2011, 30(6): 1681-1707.
[2] SAHILLIOĞLU Y. Recent Advances in Shape Correspondence. The Visual Computer, 2020, 36(8): 1705-1721.
[3] DENG B L, YAO Y X, DYKE R M, et al. A Survey of Non-Rigid 3D Registration. Computer Graphics Forum, 2022, 41(2): 559-589.
[4] COSMO L, RODOLÀ E, MASCI J, et al. Matching Deformable Objects in Clutter // Proc of the 4th International Conference on 3d Vi-sion. Washington, USA: IEEE, 2016. DOI: 10.1109/3DV.2016.10.
[5] OVSJANIKOV M, BEN-CHEN M, SOLOMON J, et al. Functional Maps: A Flexible Representation of Maps Between Shapes. ACM Transactions on Graphics, 2012, 31(4). DOI: 10.1145/2185520.2185526.
[6] MELZI S, RODOLÀ E, CASTELLANI U, et al. Localized Manifold Harmonics for Spectral Shape Analysis. Computer Graphics Forum, 2018, 37(6): 20-34.
[7] MANDAD M, COHEN-STEINER D, KOBBELT L, et al. Variance-Minimizing Transport Plans for Inter-Surface Mapping. ACM Tran-sactions on Graphics, 2017, 36(4). DOI: 10.1145/3072959.3073671.
[8] EZUZ D, SOLOMON J, BEN-CHEN M.Reversible Harmonic Maps between Discrete Surfaces. ACM Transactions on Graphics, 2019, 38(2). DOI: 10.1145/3202660.
[9] EDELSTEIN M, EZUZ D, BEN-CHEN M.ENIGMA: Evolutionary Non-Isometric Geometry Matching. ACM Transactions on Graphics,2020, 39(4). DOI: 10.1145/3386569.3392447.
[10] EISENBERGER M, LAHNER Z, CREMERS D.Smooth Shells: Multi-scale Shape Registration with Functional Maps // Proc of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2020: 12262-12271.
[11] PANINE M, KIRGO M, OVSJANIKOV M.Non-Isometric Shape Matching via Functional Maps on Landmark-Adapted Bases. Computer Graphics Forum, 2022, 41(6): 394-417.
[12] SAHILLIOĞLU Y, YEMEZ Y. Multiple Shape Correspondence by Dynamic Programming. Computer Graphics Forum, 2014, 33(7): 121-130.
[13] 杨军,雷鸣.结合函数映射与循环一致性约束的模型簇对应关系计算.激光与光电子学进展, 2019, 56(8): 123-133.
(YANG J, LEI M.Correspondence Calculation of Model Cluster by Functional Mapping combined with Cycle-Consistency Constraints. Laser and Optoelectronics Progress, 2019, 56(8): 123-133.)
[14] HUANG R Q, REN J, WONKA P, et al. CONSISTENT ZOOM-OUT: Efficient Spectral Map Synchronization. Computer Graphics Forum, 2020, 39(5): 265-278.
[15] MELZI S, REN J, RODOLÀ E, et al. ZOOMOUT: Spectral Upsampling for Efficient Shape Correspondence. ACM Transactions on Graphics, 2019, 38(6). DOI: 10.1145/3355089.3356524.
[16] HUANG R Q, ACHLIOPTAS P, GUIBAS L, et al. Limit Shapes-A Tool for Understanding Shape Differences and Variability in 3D Model Collections. Computer Graphics Forum, 2019, 38(5): 187-202.
[17] GAO M L, LAHNER Z, THUNBERG J, et al. Isometric Multi-Shape Matching // Proc of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2021: 14178-14188.
[18] 杨军,李金泰,高志明.无监督的三维模型簇对应关系协同计算.浙江大学学报(工学版), 2022, 56(10): 1935-1947.
(YANG J, LI J T, GAO Z M.Unsupervised Co-calculation on Correspondence of Three-Dimensional Shape Collections. Journal of Zhejiang University(Engineering Science), 2022, 56(10): 1935-1947.)
[19] CAO D L, BERNARD F.Unsupervised Deep Multi-Shape Mat-ching // Proc of the European Conference on Computer Vision. Berlin, Germany: Springer, 2022: 55-71.
[20] LITANY O, RODOLÀ E, BRONSTEIN A M, et al. Non-Rigid Puzzles. Computer Graphics Forum, 2016, 35(5): 135-143.
[21] WU Y, YANG J.Multi-part Shape Matching by Simultaneous Par-tial Functional Correspondence. The Visual Computer, 2023, 39(6): 393-412.
[22] SHARP N, ATTAIKI S, CRANE K,et al. DiffusionNet: Discretization Agnostic Learning on Surfaces. ACM Transactions on Gra-phics, 2022, 41(3). DOI: 10.1145/3507905.
[23] MYRONENKO A, SONG X B.Point Set Registration: Coherent Point Drift. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2010, 32(12): 2262-2275.
[24] 段凡丁. 关于最短路径的 SPFA 快速算法.西南交通大学学报, 1994, 29(2): 207-212.
(DUAN F D.A Faster Algorithm for Shortest-Path-SPFA. Journal of Southwest Jiaotong University, 1994, 29(2): 207-212.)
[25] 吴衍,杨军.内外特征一致性对齐的非等距三维模型对应关系计算.计算机辅助设计与图形学学报, 2023, 35(5): 749-759.
(WU Y, YANG J.Correspondence Calculation of Non-Isometric 3D Shapes by Intrinsic-Extrinsic Feature Alignment. Journal of Computer-Aided Design and Computer Graphics, 2023, 35(5): 749-759.)
[26] SORKINE O, ALEXA M.As-Rigid-As-Possible Surface Modeling[C/OL]. [2023-11-09].https://www.cs.jhu.edu/~misha/ReadingSeminar/Papers/Sorkine07.pdf.
[27] KOVNATSKY A, GLASHOFF K, BRONSTEIN M M.MADMM: A Generic Algorithm for Non-Smooth Optimization on Manifolds // Proc of the European Conference on Computer Vision. Berlin, Germany: Springer, 2016: 680-696.
[28] SUN M Z, MAO S W, JIANG P H, et al. Spatially and Spectrally Consistent Deep Functional Maps // Proc of the IEEE/CVF International Conference on Computer Vision. Washington, USA: IEEE, 2023: 14451-14561.
[29] MAGNET R, OVSJANIKOV M.Scalable and Efficient Functional Map Computations on Dense Meshes. Computer Graphics Forum, 2023, 42(2): 89-101.
[30] ZUFFI S, KANAZAWA A, JACOBS D W, et al. 3D Menagerie: Modeling the 3D Shape and Pose of Animals // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2017: 5524-5532.
[31] MAGNET R, REN J, SORKINE-HORNUNG O, et al. Smooth Non-Rigid Shape Matching via Effective Dirichlet Energy Optimization // Proc of the International Conference on 3D Vision. Washington, USA: IEEE, 2022: 495-504.
[32] LÄHNER Z, RODOLÀ E, BRONSTEIN M M, et al. SHRCE'16: Matching of Deformable Shapes with Topological Noise // Proc of the Eurographics Workshop on 3D Object Retrieval. Berlin, Germany: Springer, 2016: 55-60.
[33] MELZI S, MARIN R, RODOLÀ E, et al. SHREC 2019: Mat-ching Humans with Different Connectivity // Proc of the Eurographics Workshop on 3D Object Retrieval. Berlin, Germany: Springer, 2019: 121-128.
[34] BOGO F, ROMERO J, LOPER M, et al. FAUST: Dataset and Eva-luation for 3D Mesh Registration // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2014: 3794-3801.
[35] ANGUELOV D, SRINIVASAN P, KOLLER D, et al. SCAPE: Shape Completion and Animation of People. ACM Transactions on Graphics, 2005, 24(3): 408-416.
[36] LI Y, TAKEHARA H, TAKETOMI T, et al. 4DComplete: Non-Rigid Motion Estimation beyond the Observable Surface // Proc of the IEEE/CVF International Conference on Computer Vision. Washington, USA: IEEE, 2021: 12686-12696.
[37] KIM V G, LIPMAN Y, FUNKHOUSER T.3 Blended Intrinsic Maps. ACM Transactions on Graphics, 2011, 30(4). DOI: 10.1145/2010324.1964974.
[38] REN J, POULENARD A, WONKA P, et al. Continuous and Orien-tation-Preserving Correspondences via Functional Maps. ACM Transactions on Graphics, 2018, 37(6). DOI: 10.1145/3272127.3275040.