3DPCTN: Two 3D Local-Object Point-Cloud-Completion Transformer Networks Based on Self-Attention and Multi-Resolution
Acquiring high-fidelity 3D models from real-world scans is challenging. Existing shape completion-methods are incapable of generating details of objects or learning complex point dis-tributions. To address this problem, we propose two transformer-based point-cloud-completion networks and a coarse-to...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI
2022
|
| Subjects: | |
| Online Access: | View Fulltext in Publisher |