MU-GeNeRF: Multi-view Uncertainty-guided Generalizable Neural Radiance Fields for Distractor-aware Scene

TL;DR

MU-GeNeRF introduces a multi-view uncertainty-guided framework to improve neural radiance fields' robustness against transient distractors, enhancing scene reconstruction quality in challenging real-world scenarios.

Contribution

It proposes a novel uncertainty decomposition approach with a heteroscedastic loss to effectively suppress distractors and improve generalizable NeRF performance.

Findings

Outperforms existing GeNeRF methods in distractor scenarios.

Achieves comparable results to scene-specific distractor-free NeRFs.

Effectively suppresses transient distractors through uncertainty-guided supervision.

Abstract

Generalizable Neural Radiance Fields (GeNeRFs) enable high-quality scene reconstruction from sparse views and can generalize to unseen scenes. However, in real-world settings, transient distractors break cross-view structural consistency, corrupting supervision and degrading reconstruction quality. Existing distractor-free NeRF methods rely on per-scene optimization and estimate uncertainty from per-view reconstruction errors, which are not reliable for GeNeRFs and often misjudge inconsistent static structures as distractors. To this end, we propose MU-GeNeRF, a Multi-view Uncertainty-guided distractor-aware GeNeRF framework designed to alleviate GeNeRF's robust modeling challenges in the presence of transient distractions. We decompose distractor awareness into two complementary uncertainty components: Source-view Uncertainty, which captures structural discrepancies across source views caused by viewpoint changes or dynamic factors; and Target-view Uncertainty, which detects observation anomalies in the target image induced by transient distractors.These two uncertainties address distinct error sources and are combined through a heteroscedastic reconstruction loss, which guides the model to adaptively modulate supervision, enabling more robust distractor suppression and geometric modeling.Extensive experiments show that our method not only surpasses existing GeNeRFs but also achieves performance comparable to scene-specific distractor-free NeRFs.