NeRFDeformer: NeRF Transformation from a Single View via 3D Scene Flows

TL;DR

NeRFDeformer enables the modification of neural radiance fields from a single view by modeling scene transformations as 3D flows, using a novel correspondence algorithm and a new dataset for evaluation.

Contribution

The paper introduces a novel method for transforming NeRFs from a single view using 3D scene flows and a new correspondence algorithm, along with a dedicated dataset.

Findings

Outperforms existing NeRF editing methods

Effective filtering of false correspondences

Robust single-view scene modification

Abstract

We present a method for automatically modifying a NeRF representation based on a single observation of a non-rigid transformed version of the original scene. Our method defines the transformation as a 3D flow, specifically as a weighted linear blending of rigid transformations of 3D anchor points that are defined on the surface of the scene. In order to identify anchor points, we introduce a novel correspondence algorithm that first matches RGB-based pairs, then leverages multi-view information and 3D reprojection to robustly filter false positives in two steps. We also introduce a new dataset for exploring the problem of modifying a NeRF scene through a single observation. Our dataset ( https://github.com/nerfdeformer/nerfdeformer ) contains 113 synthetic scenes leveraging 47 3D assets. We show that our proposed method outperforms NeRF editing methods as well as diffusion-based methods, and we also explore different methods for filtering correspondences.