WTFM Layer: An Effective Map Extractor for Unsupervised Shape Correspondence
文献类型：期刊论文
| 作者 | Liu, Shengjun2; Xu, Haojun2; Yan, Dong-Ming3 ; Hu, Ling1; Liu, Xinru2; Li, Qinsong2,4
|
| 刊名 | COMPUTER GRAPHICS FORUM
 |
| 出版日期 | 2022-10-01 |
| 卷号 | 41期号:7页码:51-61 |
| ISSN号 | 0167-7055 |
| DOI | 10.1111/cgf.14656 |
| 通讯作者 | Li, Qinsong(qinsli.cg@foxmail.com) |
| 英文摘要 | We propose a novel unsupervised learning approach for computing correspondences between non-rigid 3D shapes. The core idea is that we integrate a novel structural constraint into the deep functional map pipeline, a recently dominant learning framework for shape correspondence, via a powerful spectral manifold wavelet transform (SMWT). As SMWT is isometrically invariant operator and can analyze features from multiple frequency bands, we use the multiscale SMWT results of the learned features as function preservation constraints to optimize the functional map by assuming each frequency-band information of the descriptors should be correspondingly preserved by the functional map. Such a strategy allows extracting significantly more deep feature information than existing approaches which only use the learned descriptors to estimate the functional map. And our formula strongly ensure the isometric properties of the underlying map. We also prove that our computation of the functional map amounts to filtering processes only referring to matrix multiplication. Then, we leverage the alignment errors of intrinsic embedding between shapes as a loss function and solve it in an unsupervised way using the Sinkhorn algorithm. Finally, we utilize DiffusionNet as a feature extractor to ensure that discretization-resistant and directional shape features are produced. Experiments on multiple challenging datasets prove that our method can achieve state-of-the-art correspondence quality. Furthermore, our method yields significant improvements in robustness to shape discretization and generalization across the different datasets. The source code and trained models will be available at https://github.com/HJ-Xu/WTFM-Layer. |
| 资助项目 | Natural Science Foundation of China[62172447] ; Natural Science Foundation of China[61876191] ; Natural Science Foundation of China[62172415] ; Hunan Provincial Natural Science Foundation of China[2021JJ30172] ; Open Project Program of the National Laboratory of Pattern Recognition (NLPR)[202200025] ; Fundamental Research Funds for the Central Universities of Central South University[2022ZZTS0606] |
| WOS研究方向 | Computer Science |
| 语种 | 英语 |
| 出版者 | WILEY |
| WOS记录号 | WOS:001008797000006 |
| 资助机构 | Natural Science Foundation of China ; Hunan Provincial Natural Science Foundation of China ; Open Project Program of the National Laboratory of Pattern Recognition (NLPR) ; Fundamental Research Funds for the Central Universities of Central South University |
| 源URL | [http://ir.ia.ac.cn/handle/173211/53522]  |
| 专题 | 多模态人工智能系统全国重点实验室 |
| 通讯作者 | Li, Qinsong |
| 作者单位 | 1.Hunan First Normal Univ, Sch Math & Stat, Changsha 410205, Hunan, Peoples R China 2.Cent South Univ, Inst Engn Modeling & Sci Comp, Changsha 410083, Hunan, Peoples R China 3.Chinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China 4.Cent South Univ, Big Data Inst, Changsha 410083, Hunan, Peoples R China |
| 推荐引用方式 GB/T 7714 | Liu, Shengjun,Xu, Haojun,Yan, Dong-Ming,et al. WTFM Layer: An Effective Map Extractor for Unsupervised Shape Correspondence[J]. COMPUTER GRAPHICS FORUM,2022,41(7):51-61. |
| APA | Liu, Shengjun,Xu, Haojun,Yan, Dong-Ming,Hu, Ling,Liu, Xinru,&Li, Qinsong.(2022).WTFM Layer: An Effective Map Extractor for Unsupervised Shape Correspondence.COMPUTER GRAPHICS FORUM,41(7),51-61. |
| MLA | Liu, Shengjun,et al."WTFM Layer: An Effective Map Extractor for Unsupervised Shape Correspondence".COMPUTER GRAPHICS FORUM 41.7(2022):51-61. |
入库方式： OAI收割
来源：自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。