AuraFusion360: Augmented Unseen Region Alignment for Reference-based 360{\deg} Unbounded Scene Inpainting

TL;DR

AuraFusion360 is a novel 3D scene inpainting method that achieves high-quality, view-consistent results in 360-degree unbounded scenes by combining depth-aware masking, zero-shot depth diffusion, and detail enhancement.

Contribution

It introduces a comprehensive framework with novel depth-aware masking, zero-shot depth diffusion, and a new dataset for 360-degree scene inpainting, advancing the state of the art.

Findings

Outperforms existing methods in perceptual quality

Maintains geometric accuracy across viewpoints

Introduces the first dataset for 360-degree scene inpainting

Abstract

Three-dimensional scene inpainting is crucial for applications from virtual reality to architectural visualization, yet existing methods struggle with view consistency and geometric accuracy in 360{\deg} unbounded scenes. We present AuraFusion360, a novel reference-based method that enables high-quality object removal and hole filling in 3D scenes represented by Gaussian Splatting. Our approach introduces (1) depth-aware unseen mask generation for accurate occlusion identification, (2) Adaptive Guided Depth Diffusion, a zero-shot method for accurate initial point placement without requiring additional training, and (3) SDEdit-based detail enhancement for multi-view coherence. We also introduce 360-USID, the first comprehensive dataset for 360{\deg} unbounded scene inpainting with ground truth. Extensive experiments demonstrate that AuraFusion360 significantly outperforms existing methods, achieving superior perceptual quality while maintaining geometric accuracy across dramatic viewpoint changes.