PanoPlane: Plane-Aware Panoramic Completion for Sparse-View Indoor 3D Gaussian Splatting

TL;DR

PanoPlane is a novel method for indoor 3D scene reconstruction and view synthesis that uses panoramic completion and attention steering to improve quality from sparse views without additional training.

Contribution

It introduces Layout Anchored Attention Steering, a training-free technique that guides diffusion models to focus on planar surfaces for better scene completion.

Findings

Achieves up to +17.8% PSNR improvement over state-of-the-art.

Enables accurate novel view synthesis from as few as three input views.

Demonstrates superior performance on Replica, ScanNet++, and Matterport3D datasets.

Abstract

We present PanoPlane, an approach for high-fidelity sparse-view indoor novel view synthesis that reconstructs closed room geometry via panoramic scene completion. Unlike perspective-based methods that generate training views from limited fields of view, PanoPlane leverages $360^{\circ}$ panoramic completion to condition the generative process on the full spatial layout. We propose Layout Anchored Attention Steering, a training-free mechanism that steers attention within the diffusion model's internal representation toward scene's detected planar surfaces at inference time. By directing each unobserved region's attention toward geometrically consistent observed content, our method replaces unconstrained hallucination with grounded surface extrapolation. The resulting panoramic completions provide supervision for 3D Gaussian Splatting, enabling accurate novel-view synthesis across unobserved regions from as few as three input views. Experiments on Replica, ScanNet++, and Matterport3D demonstrate state-of-the-art novel view synthesis quality across 3, 6, and 9 input views, achieving up to $+17.8\%$ improvement in PSNR over the current state-of-the-art baseline without any training or fine-tuning of the diffusion model.