AMS-Net: An Attention-Based Multi-Scale Network for Classification of 3D Terracotta Warrior Fragments
As an essential step in the restoration of Terracotta Warriors, the results of fragments classification will directly affect the performance of fragments matching and splicing. However, most of the existing methods are based on traditional technology and have low accuracy in classification. A practi...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-09-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/13/18/3713 |
| _version_ | 1797517334264414208 |
|---|---|
| author | Jie Liu Xin Cao Pingchuan Zhang Xueli Xu Yangyang Liu Guohua Geng Fengjun Zhao Kang Li Mingquan Zhou |
| author_facet | Jie Liu Xin Cao Pingchuan Zhang Xueli Xu Yangyang Liu Guohua Geng Fengjun Zhao Kang Li Mingquan Zhou |
| author_sort | Jie Liu |
| collection | DOAJ |
| description | As an essential step in the restoration of Terracotta Warriors, the results of fragments classification will directly affect the performance of fragments matching and splicing. However, most of the existing methods are based on traditional technology and have low accuracy in classification. A practical and effective classification method for fragments is an urgent need. In this case, an attention-based multi-scale neural network named AMS-Net is proposed to extract significant geometric and semantic features. AMS-Net is a hierarchical structure consisting of a multi-scale set abstraction block (MS-BLOCK) and a fully connected (FC) layer. MS-BLOCK consists of a local-global layer (LGLayer) and an improved multi-layer perceptron (IMLP). With a multi-scale strategy, LGLayer can parallel extract the local and global features from different scales. IMLP can concatenate the high-level and low-level features for classification tasks. Extensive experiments on the public data set (ModelNet40/10) and the real-world Terracotta Warrior fragments data set are conducted. The accuracy results with normal can achieve 93.52% and 96.22%, respectively. For real-world data sets, the accuracy is best among the existing methods. The robustness and effectiveness of the performance on the task of 3D point cloud classification are also investigated. It proves that the proposed end-to-end learning network is more effective and suitable for the classification of the Terracotta Warrior fragments. |
| first_indexed | 2024-03-10T07:15:14Z |
| format | Article |
| id | doaj.art-33fff80e17434ee09365515c6c84dc7c |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-10T07:15:14Z |
| publishDate | 2021-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-33fff80e17434ee09365515c6c84dc7c2023-11-22T15:07:14ZengMDPI AGRemote Sensing2072-42922021-09-011318371310.3390/rs13183713AMS-Net: An Attention-Based Multi-Scale Network for Classification of 3D Terracotta Warrior FragmentsJie Liu0Xin Cao1Pingchuan Zhang2Xueli Xu3Yangyang Liu4Guohua Geng5Fengjun Zhao6Kang Li7Mingquan Zhou8School of Information Science and Technology, Northwest University, Xi’an 710068, ChinaSchool of Information Science and Technology, Northwest University, Xi’an 710068, ChinaSchool of Information Engineering, Henan Institute of Science and Technology, Xinxiang 453003, ChinaSchool of Information Science and Technology, Northwest University, Xi’an 710068, ChinaSchool of Information Science and Technology, Northwest University, Xi’an 710068, ChinaSchool of Information Science and Technology, Northwest University, Xi’an 710068, ChinaSchool of Information Science and Technology, Northwest University, Xi’an 710068, ChinaSchool of Information Science and Technology, Northwest University, Xi’an 710068, ChinaSchool of Information Science and Technology, Northwest University, Xi’an 710068, ChinaAs an essential step in the restoration of Terracotta Warriors, the results of fragments classification will directly affect the performance of fragments matching and splicing. However, most of the existing methods are based on traditional technology and have low accuracy in classification. A practical and effective classification method for fragments is an urgent need. In this case, an attention-based multi-scale neural network named AMS-Net is proposed to extract significant geometric and semantic features. AMS-Net is a hierarchical structure consisting of a multi-scale set abstraction block (MS-BLOCK) and a fully connected (FC) layer. MS-BLOCK consists of a local-global layer (LGLayer) and an improved multi-layer perceptron (IMLP). With a multi-scale strategy, LGLayer can parallel extract the local and global features from different scales. IMLP can concatenate the high-level and low-level features for classification tasks. Extensive experiments on the public data set (ModelNet40/10) and the real-world Terracotta Warrior fragments data set are conducted. The accuracy results with normal can achieve 93.52% and 96.22%, respectively. For real-world data sets, the accuracy is best among the existing methods. The robustness and effectiveness of the performance on the task of 3D point cloud classification are also investigated. It proves that the proposed end-to-end learning network is more effective and suitable for the classification of the Terracotta Warrior fragments.https://www.mdpi.com/2072-4292/13/18/3713self-attentionmulti-scaledeep neural networkspoint cloud classificationTerracotta Warrior fragments |
| spellingShingle | Jie Liu Xin Cao Pingchuan Zhang Xueli Xu Yangyang Liu Guohua Geng Fengjun Zhao Kang Li Mingquan Zhou AMS-Net: An Attention-Based Multi-Scale Network for Classification of 3D Terracotta Warrior Fragments Remote Sensing self-attention multi-scale deep neural networks point cloud classification Terracotta Warrior fragments |
| title | AMS-Net: An Attention-Based Multi-Scale Network for Classification of 3D Terracotta Warrior Fragments |
| title_full | AMS-Net: An Attention-Based Multi-Scale Network for Classification of 3D Terracotta Warrior Fragments |
| title_fullStr | AMS-Net: An Attention-Based Multi-Scale Network for Classification of 3D Terracotta Warrior Fragments |
| title_full_unstemmed | AMS-Net: An Attention-Based Multi-Scale Network for Classification of 3D Terracotta Warrior Fragments |
| title_short | AMS-Net: An Attention-Based Multi-Scale Network for Classification of 3D Terracotta Warrior Fragments |
| title_sort | ams net an attention based multi scale network for classification of 3d terracotta warrior fragments |
| topic | self-attention multi-scale deep neural networks point cloud classification Terracotta Warrior fragments |
| url | https://www.mdpi.com/2072-4292/13/18/3713 |
| work_keys_str_mv | AT jieliu amsnetanattentionbasedmultiscalenetworkforclassificationof3dterracottawarriorfragments AT xincao amsnetanattentionbasedmultiscalenetworkforclassificationof3dterracottawarriorfragments AT pingchuanzhang amsnetanattentionbasedmultiscalenetworkforclassificationof3dterracottawarriorfragments AT xuelixu amsnetanattentionbasedmultiscalenetworkforclassificationof3dterracottawarriorfragments AT yangyangliu amsnetanattentionbasedmultiscalenetworkforclassificationof3dterracottawarriorfragments AT guohuageng amsnetanattentionbasedmultiscalenetworkforclassificationof3dterracottawarriorfragments AT fengjunzhao amsnetanattentionbasedmultiscalenetworkforclassificationof3dterracottawarriorfragments AT kangli amsnetanattentionbasedmultiscalenetworkforclassificationof3dterracottawarriorfragments AT mingquanzhou amsnetanattentionbasedmultiscalenetworkforclassificationof3dterracottawarriorfragments |