We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CV

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Computer Vision and Pattern Recognition

Title: Spline Positional Encoding for Learning 3D Implicit Signed Distance Fields

Abstract: Multilayer perceptrons (MLPs) have been successfully used to represent 3D shapes implicitly and compactly, by mapping 3D coordinates to the corresponding signed distance values or occupancy values. In this paper, we propose a novel positional encoding scheme, called Spline Positional Encoding, to map the input coordinates to a high dimensional space before passing them to MLPs, for helping to recover 3D signed distance fields with fine-scale geometric details from unorganized 3D point clouds. We verified the superiority of our approach over other positional encoding schemes on tasks of 3D shape reconstruction from input point clouds and shape space learning. The efficacy of our approach extended to image reconstruction is also demonstrated and evaluated.
Comments: Accepted by IJCAI 2021; Code: this https URL
Subjects: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR)
Cite as: arXiv:2106.01553 [cs.CV]
  (or arXiv:2106.01553v2 [cs.CV] for this version)

Submission history

From: Peng-Shuai Wang [view email]
[v1] Thu, 3 Jun 2021 02:37:47 GMT (5331kb,D)
[v2] Thu, 28 Oct 2021 06:39:04 GMT (5363kb,D)

Link back to: arXiv, form interface, contact.