Efficient 3D Shape Matching: Dense Correspondence for non-isometric Deformation
This paper presents an application of deep learning in computer graphics, utilizing learn-based networks for 3D shape matching. We propose an efficient method for shape matching between 3D models with non-isometric deformation. Our method organizes intrinsic and directional attributes in a structure...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Iran University of Science and Technology
2024-11-01
|
| Series: | Iranian Journal of Electrical and Electronic Engineering |
| Subjects: | |
| Online Access: | http://ijeee.iust.ac.ir/article-1-3504-en.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841550983434338304 |
|---|---|
| author | Amirreza Amirfathiyan Hossein Ebrahimnezhad |
| author_facet | Amirreza Amirfathiyan Hossein Ebrahimnezhad |
| author_sort | Amirreza Amirfathiyan |
| collection | DOAJ |
| description | This paper presents an application of deep learning in computer graphics, utilizing learn-based networks for 3D shape matching. We propose an efficient method for shape matching between 3D models with non-isometric deformation. Our method organizes intrinsic and directional attributes in a structured manner. For this purpose, we use a hybrid feature derived from Diffusion-Net and spectral features. In fact, we combine learned-based intrinsic properties with orientation-preserving features and demonstrate the effectiveness of our method. We achieve this by first extracting features from Diffusion-Net. Then, we compute two maps based on the functional map networks to obtain intrinsic and directional features. Finally, we combine them to achieve a desired map that can resolve symmetry ambiguities on models with high deformation. Quantitative results on the TOSCA dataset indicate that the proposed method achieves lowest average geodetic error of 0.0023, outperforming state-of-the-art methods and reducing the error by 70.66%. We demonstrate that our method outperforms similar approaches by leveraging an accurate feature extractor and effective geometric regularizers, allowing for better handling of non-isometric shapes and resulting in reduced matching errors. |
| format | Article |
| id | doaj-art-49f8b722205d4ae1987e83533f201997 |
| institution | Kabale University |
| issn | 1735-2827 2383-3890 |
| language | English |
| publishDate | 2024-11-01 |
| publisher | Iran University of Science and Technology |
| record_format | Article |
| series | Iranian Journal of Electrical and Electronic Engineering |
| spelling | doaj-art-49f8b722205d4ae1987e83533f2019972025-01-09T18:47:15ZengIran University of Science and TechnologyIranian Journal of Electrical and Electronic Engineering1735-28272383-38902024-11-01204162172Efficient 3D Shape Matching: Dense Correspondence for non-isometric DeformationAmirreza Amirfathiyan0Hossein Ebrahimnezhad1 Computer Vision Res. Lab., Electrical Engineering Faculty, Sahand University of Technology, Tabriz, Iran. Computer Vision Res. Lab., Electrical Engineering Faculty, Sahand University of Technology, Tabriz, Iran. This paper presents an application of deep learning in computer graphics, utilizing learn-based networks for 3D shape matching. We propose an efficient method for shape matching between 3D models with non-isometric deformation. Our method organizes intrinsic and directional attributes in a structured manner. For this purpose, we use a hybrid feature derived from Diffusion-Net and spectral features. In fact, we combine learned-based intrinsic properties with orientation-preserving features and demonstrate the effectiveness of our method. We achieve this by first extracting features from Diffusion-Net. Then, we compute two maps based on the functional map networks to obtain intrinsic and directional features. Finally, we combine them to achieve a desired map that can resolve symmetry ambiguities on models with high deformation. Quantitative results on the TOSCA dataset indicate that the proposed method achieves lowest average geodetic error of 0.0023, outperforming state-of-the-art methods and reducing the error by 70.66%. We demonstrate that our method outperforms similar approaches by leveraging an accurate feature extractor and effective geometric regularizers, allowing for better handling of non-isometric shapes and resulting in reduced matching errors.http://ijeee.iust.ac.ir/article-1-3504-en.pdf3d shape matching3d shape correspondenceorientation preservingdeep learning. |
| spellingShingle | Amirreza Amirfathiyan Hossein Ebrahimnezhad Efficient 3D Shape Matching: Dense Correspondence for non-isometric Deformation Iranian Journal of Electrical and Electronic Engineering 3d shape matching 3d shape correspondence orientation preserving deep learning. |
| title | Efficient 3D Shape Matching: Dense Correspondence for non-isometric Deformation |
| title_full | Efficient 3D Shape Matching: Dense Correspondence for non-isometric Deformation |
| title_fullStr | Efficient 3D Shape Matching: Dense Correspondence for non-isometric Deformation |
| title_full_unstemmed | Efficient 3D Shape Matching: Dense Correspondence for non-isometric Deformation |
| title_short | Efficient 3D Shape Matching: Dense Correspondence for non-isometric Deformation |
| title_sort | efficient 3d shape matching dense correspondence for non isometric deformation |
| topic | 3d shape matching 3d shape correspondence orientation preserving deep learning. |
| url | http://ijeee.iust.ac.ir/article-1-3504-en.pdf |
| work_keys_str_mv | AT amirrezaamirfathiyan efficient3dshapematchingdensecorrespondencefornonisometricdeformation AT hosseinebrahimnezhad efficient3dshapematchingdensecorrespondencefornonisometricdeformation |