Corrupted Point Cloud Classification Through Deep Learning with Local Feature Descriptor
Xian Wu, Xueyi Guo, Hang Peng, Bin Su, Sabbir Ahamod, Fenglin Han · 2024
Read Corrupted Point Cloud Classification Through Deep Learning with Local Feature Descriptor by Xian Wu, Xueyi Guo, Hang Peng, Bin Su, Sabbir Ahamod, Fenglin Han (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-11644878.txt.
Description
Wu, Guo, Peng, Su, Ahamod, and Han classify damaged three dimensional point clouds in twenty twenty-four. Point clouds from scanners guide self driving and face recognition systems. Real data often suffer occlusion, rotation, and noise. Tuning network structure alone did not fix the accuracy drop.
Their approach extracts local feature descriptors before classification. A new network design is built around those descriptors. Training data were augmented for ten chosen object categories. Tests covered occlusion and coordinate changes against leading models. Their model kept higher accuracy under heavy corruption. Depth camera trials on real objects confirmed the gains. The authors say feature preprocessing is the key to robustness.
Catalog record
| CREATOR | Xian Wu, Xueyi Guo, Hang Peng, Bin Su, Sabbir Ahamod, Fenglin Han |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 40 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |