FisherRF: Active View Selection and Uncertainty Quantification for Radiance Fields using Fisher Information

TL;DR

This paper introduces FisherRF, a method that uses Fisher Information to directly quantify uncertainty and select optimal views in Radiance Fields, improving efficiency and accuracy in view selection and mapping tasks.

Contribution

FisherRF is the first approach to leverage Fisher Information for direct uncertainty quantification and active view selection in Radiance Fields, outperforming existing indirect methods.

Findings

Achieves state-of-the-art results in view selection and active mapping.

Effectively quantifies uncertainty in Radiance Fields.

Demonstrates improved information gain through Fisher Information maximization.

Abstract

This study addresses the challenging problem of active view selection and uncertainty quantification within the domain of Radiance Fields. Neural Radiance Fields (NeRF) have greatly advanced image rendering and reconstruction, but the cost of acquiring images poses the need to select the most informative viewpoints efficiently. Existing approaches depend on modifying the model architecture or hypothetical perturbation field to indirectly approximate the model uncertainty. However, selecting views from indirect approximation does not guarantee optimal information gain for the model. By leveraging Fisher Information, we directly quantify observed information on the parameters of Radiance Fields and select candidate views by maximizing the Expected Information Gain(EIG). Our method achieves state-of-the-art results on multiple tasks, including view selection, active mapping, and uncertainty quantification, demonstrating its potential to advance the field of Radiance Fields.