File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Conference Paper: MAT-Net: Medial Axis Transform Network for 3D Object Recognition

TitleMAT-Net: Medial Axis Transform Network for 3D Object Recognition
Authors
KeywordsComputer Vision: 2D and 3D Computer Vision
Computer Vision: Recognition: Detection, Categorization, Indexing, Matching, Retrieval, Semantic Interpretation
Issue Date2019
PublisherInternational Joint Conference on Artificial Intelligence. The Journal's web site is located at https://www.ijcai.org/past_proceedings
Citation
Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence (IJCAI 2019), Macao, China, 10-16 August 2019, p. 774-781 How to Cite?
Abstract3D deep learning performance depends on object representation and local feature extraction. In this work, we present MAT-Net, a neural network which captures local and global features from the Medial Axis Transform (MAT). Different from K-Nearest-Neighbor method which extracts local features by a fixed number of neighbors, our MAT-Net exploits effective modules Group-MAT and Edge-Net to process topological structure. Experimental results illustrate that MAT-Net demonstrates competitive or better performance on 3D shape recognition than state-of-the-art methods, and prove that MAT representation has excellent capacity in 3D deep learning, even in the case of low resolution.
Persistent Identifierhttp://hdl.handle.net/10722/294205
ISBN

 

DC FieldValueLanguage
dc.contributor.authorHu, J-
dc.contributor.authorWang, B-
dc.contributor.authorQian, L-
dc.contributor.authorPan, Y-
dc.contributor.authorGuo, X-
dc.contributor.authorLiu, L-
dc.contributor.authorWang, WP-
dc.date.accessioned2020-11-23T08:27:55Z-
dc.date.available2020-11-23T08:27:55Z-
dc.date.issued2019-
dc.identifier.citationProceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence (IJCAI 2019), Macao, China, 10-16 August 2019, p. 774-781-
dc.identifier.isbn9780999241141-
dc.identifier.urihttp://hdl.handle.net/10722/294205-
dc.description.abstract3D deep learning performance depends on object representation and local feature extraction. In this work, we present MAT-Net, a neural network which captures local and global features from the Medial Axis Transform (MAT). Different from K-Nearest-Neighbor method which extracts local features by a fixed number of neighbors, our MAT-Net exploits effective modules Group-MAT and Edge-Net to process topological structure. Experimental results illustrate that MAT-Net demonstrates competitive or better performance on 3D shape recognition than state-of-the-art methods, and prove that MAT representation has excellent capacity in 3D deep learning, even in the case of low resolution.-
dc.languageeng-
dc.publisherInternational Joint Conference on Artificial Intelligence. The Journal's web site is located at https://www.ijcai.org/past_proceedings-
dc.relation.ispartofProceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence (IJCAI 2019)-
dc.subjectComputer Vision: 2D and 3D Computer Vision-
dc.subjectComputer Vision: Recognition: Detection, Categorization, Indexing, Matching, Retrieval, Semantic Interpretation-
dc.titleMAT-Net: Medial Axis Transform Network for 3D Object Recognition-
dc.typeConference_Paper-
dc.identifier.emailWang, WP: wenping@cs.hku.hk-
dc.identifier.authorityWang, WP=rp00186-
dc.description.naturelink_to_OA_fulltext-
dc.identifier.doi10.24963/ijcai.2019/109-
dc.identifier.hkuros319285-
dc.identifier.spage774-
dc.identifier.epage781-
dc.publisher.placeUnited States-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats