Point cloud upsampling using deep self-sampling with point saliency

The paper has been accepted by Journal of Mechanical Science and Technology (JMST) 2023 (Q2 SCIE)
"Point cloud upsampling using deep self-sampling with point saliency".

Abstract

Point cloud upsampling is a process of increasing the point density to represent an object or environment effectively. Recent studies have focused on deep learning-based approaches that learn mapping from a sparse to a dense region of the point cloud. Self-supervised learning-based upsampling techniques have gained attention due to their capability to learn predefined characteristics without previous training on a large dataset. This study proposes deep self-sampling with point saliency. The approach involves the use of a self-sampling network with two predefined consolidation strategies, namely density and curvature, along with a saliency feature, to restore the underlying characteristics of an object effectively. Additionally, multistep upsampling is applied to determine the best order of different consolidation strategies for optimal results. Experimental results show that multistep self-sampling using point saliency outperforms the existing approach because it can effectively restore the underlying shapes of the object qualitatively and quantitatively.

---

This repository is not the official implementation of the paper.

Instead, this repository serves as a reference point. The actual implementation and extensions appear in our follow-up study, which builds upon this paper.

Saliency Code Reference

Although this repository contains no code, the saliency module from the original paper is implemented in our follow-up study:

• File: Denoise-yourself/data_handler.py
• Mode: saliency

Access the full follow-up repository here: Denoise-yourself
Specific saliency implementation: data_handler.py (saliency mode)

---

Citation

@article{hur2023point,
  title={Point cloud upsampling using deep self-sampling with point saliency},
  author={Hur, Ji-Hyeon and Kim, Hyungki and Kwon, Soonjo},
  journal={Journal of Mechanical Science and Technology},
  volume={37},
  number={12},
  pages={6083--6091},
  year={2023},
  publisher={Springer}
}