TU Darmstadt / ULB / TUprints

A Post Processing Technique to Automatically Remove Floater Artifacts in Neural Radiance Fields

Wirth, T. ; Rak, A. ; Knauthe, V. ; Fellner, D. W. (2024)
A Post Processing Technique to Automatically Remove Floater Artifacts in Neural Radiance Fields.
In: Computer Graphics Forum, 2023, 42 (7)
doi: 10.26083/tuprints-00027236
Article, Secondary publication, Publisher's Version

[img] Text
Copyright Information: CC BY 4.0 International - Creative Commons, Attribution.

Download (22MB)
Item Type: Article
Type of entry: Secondary publication
Title: A Post Processing Technique to Automatically Remove Floater Artifacts in Neural Radiance Fields
Language: English
Date: 27 May 2024
Place of Publication: Darmstadt
Year of primary publication: October 2023
Place of primary publication: Oxford
Publisher: Wiley-Blackwell
Journal or Publication Title: Computer Graphics Forum
Volume of the journal: 42
Issue Number: 7
Collation: 12 Seiten
DOI: 10.26083/tuprints-00027236
Corresponding Links:
Origin: Secondary publication DeepGreen

Neural Radiance Fields have revolutionized Novel View Synthesis by providing impressive levels of realism. However, in most in‐the‐wild scenes they suffer from floater artifacts that occur due to sparse input images or strong view‐dependent effects. We propose an approach that uses neighborhood based clustering and a consistency metric on NeRF models trained on different scene scales to identify regions that contain floater artifacts based on Instant‐NGPs multiscale occupancy grids. These occupancy grids contain the position of relevant optical densities in the scene. By pruning the regions that we identified as containing floater artifacts, they are omitted during the rendering process, leading to higher quality resulting images. Our approach has no negative runtime implications for the rendering process and does not require retraining of the underlying Multi Layer Perceptron. We show on a qualitative base, that our approach is suited to remove floater artifacts while preserving most of the scenes relevant geometry. Furthermore, we conduct a comparison to state‐of‐the‐art techniques on the Nerfbusters dataset, that was created with measuring the implications of floater artifacts in mind. This comparison shows, that our method outperforms currently available techniques. Our approach does not require additional user input, but can be be used in an interactive manner. In general, the presented approach is applicable to every architecture that uses an explicit representation of a scene's occupancy distribution to accelerate the rendering process.

Identification Number: Artikel-ID: e14977
Status: Publisher's Version
URN: urn:nbn:de:tuda-tuprints-272362
Classification DDC: 000 Generalities, computers, information > 004 Computer science
Divisions: 20 Department of Computer Science > Interactive Graphics Systems
20 Department of Computer Science > Fraunhofer IGD
Date Deposited: 27 May 2024 12:55
Last Modified: 27 May 2024 12:55
SWORD Depositor: Deep Green
URI: https://tuprints.ulb.tu-darmstadt.de/id/eprint/27236
Actions (login required)
View Item View Item