Active Sampling and Gaussian Reconstruction for Radio Frequency Radiance Field

Chi-Shiang Gau, Xingyu Chen, Tara Javidi, Xinyu Zhang · 2024

Read Active Sampling and Gaussian Reconstruction for Radio Frequency Radiance Field by Chi-Shiang Gau, Xingyu Chen, Tara Javidi, Xinyu Zhang (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/arxiv-2412-08003.txt.

Description

Chi-Shiang Gau and co-authors rebuild radio signal maps with a Gaussian method that needs no training. Their 2024 paper deals with radio frequency radiance fields, which describe how signal strength varies through a space. Reflections and diffraction make such fields hard to model, especially when objects are unknown. A neural approach worked but needed many samples and heavy computing. It also gave no estimate of its own uncertainty.

The new method fits the field with Gaussians directly. It reaches the same quality with far fewer samples. It also reports a confidence value for any point in the scene. Adding active sampling, which picks the most useful next measurement spots, cuts the sample count further. In changing scenes, the method adapts without restarting the whole process. The authors present it as a lighter path to signal mapping.

Catalog record

CREATORChi-Shiang Gau, Xingyu Chen, Tara Javidi, Xinyu Zhang
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE32 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-08003

Browse the full library · About the Tower of Babel