MACRO: Training-free Multi-plane Attention for Closeup Render Optimization

2026-07-04

AI Summary

This work addresses the degradation in rendering quality of 3D Gaussian Splatting (3DGS) for close-up views, which arises from scale discrepancies between training and target viewpoints, as well as the significant artifacts introduced by existing diffusion-based methods. To overcome these limitations, the authors propose a training-free optimization approach that leverages the known 3D scene structure to decompose close-up views into depth planes. By cropping and scale-aligning reference views in image space and introducing a depth-aware multi-plane cross-attention mechanism, the method ensures accurate cross-scale feature correspondence. The study further reveals, for the first time, that the lack of scale invariance in reference-conditioned diffusion models is the fundamental cause of close-up rendering failure. Evaluated on two newly established benchmarks for close-up novel view synthesis, the proposed method outperforms current 3DGS and diffusion-based approaches in both reconstruction accuracy and perceptual quality.

Abstract

Close-up rendering, zooming into a scene well beyond any training camera, is important for virtual production and interactive 3D content, yet remains an open challenge. 3D Gaussian splatting (3DGS) enables high-fidelity, real-time novel view synthesis, but its rendering quality degrades at close range. Recent diffusion-based methods that enhance the rendering by conditioning on reference images from the training set produce significant artifacts in this setting. We analyze this failure and identify its root cause: the scale gap between the close-up and reference views. We show that the features in reference-conditioned enhancement models are not scale-invariant, causing cross-view attention to retrieve incorrect correspondences when the same content appears at different scales, and that this mismatch cannot be corrected in latent space because the VAE encoder is not scale-equivariant. Building on this analysis we introduce MACRO, Multi-plane Attention for Closeup Render Optimization, a training-free method for high-quality close-up novel view synthesis from 3DGS. MACRO resolves the scale gap by leveraging the scene's known 3D structure: it decomposes the close-up into depth planes, crops and resizes references in image space to match the scale of each plane before encoding, and applies a depth-aware attention mask so each token attends only to scale-matched references. The method requires no architectural changes or additional training. We further contribute two new close-up novel view synthesis benchmarks, the first standardized evaluation protocol for this setting, and demonstrate state-of-the-art results on both, outperforming existing 3DGS and diffusion-based methods on both reconstruction and perceptual metrics. Project page: https://nitzanhod.github.io/MACRO

Problem

• close-up rendering
• scale gap
• novel view synthesis
• 3D Gaussian splatting
• cross-view attention

Innovation

• Multi-plane Attention
• Close-up Rendering
• Scale-invariant Correspondence
• 3D Gaussian Splatting
• Training-free Enhancement