4D-Editor: Interactive Object-level Editing in Dynamic Neural Radiance Fields via Semantic Distillation

Remove woman

Remove balloon

Shift + Recolor + Remove



Filter balloon

Remove balloon

Composite



Comparison with Baselines

N3F

ISRF

Ours

Remove Balloon



N3F

ISRF

Ours

Remove Balloon



N3F

ISRF

Ours

Remove Woman



GUI

Abstract

This paper targets interactive object-level editing (e.g., deletion, recoloring, transformation, composition) in dynamic scenes. Recently, some methods aiming for flexible editing static scenes represented by neural radiance field (NeRF) have shown impressive synthesis quality, while similar capabilities in time-variant dynamic scenes remain limited. To solve this problem, we propose 4D-Editor, an interactive semantic-driven editing framework, allowing editing multiple objects in a dynamic NeRF with user strokes on a single frame. We propose an extension to the original dynamic NeRF by incorporating a hybrid semantic feature distillation to maintain spatial-temporal consistency after editing. In addition, we design Recursive Selection Refinement that significantly boosts object segmentation accuracy within a dynamic NeRF to aid the editing process. Moreover, we develop Multi-view Reprojection Inpainting to fill holes caused by incomplete scene capture after editing. Extensive experiments and editing examples on real-world demonstrate that 4D-Editor achieves photo-realistic editing on dynamic NeRFs.


Object-level Editing on Multiple Objects

Origin: Stroller

Origin: Breakdance



Remove adult

Remove boy + Recolor board



More Editing Operations

Composite: Depth

Composite

Timing Reversal



Hyalinize

Mirror

Scale



Hybrid Semantic Radiance Field Reconstruction

DINO Semantic Features (Coarse)



After Semantic Features Distillation (Fine)



Recolor in HSL Structure

<- Hue

Saturation +

Lightness +



Hue ->

Saturation -

Lightness -