Resonant Brane Splatting for Arbitrary-Scale Super-Resolution

2026-06-28

AI Summary

This work addresses the limitations of conventional Gaussian splatting in arbitrary-scale super-resolution (ASR), which struggles to efficiently represent sharp edges and fine textures due to its reliance on smooth, low-pass basis functions and suffers from rasterization bottlenecks. To overcome these issues, the authors propose a Resonant Blob Splatting (RBS) framework that replaces Gaussian primitives with spatially varying colored blobs embedded with Gaussian-Hermite modes, enabling single-footprint modeling of local contrast and intricate structures. Furthermore, they introduce a differentiable rasterizer grounded in quantum turning points and an accurate culling strategy, substantially reducing computational overhead. Evaluated on standard ASR benchmarks, RBS outperforms existing implicit and Gaussian splatting approaches, achieving a superior trade-off between reconstruction fidelity and inference efficiency.

Abstract

Arbitrary-Scale Super-Resolution (ASR) reconstructs images at continuous magnification factors. Recent methods accelerate inference by replacing computationally heavy implicit neural decoders with explicit 2D Gaussian Splatting (GS). However, since standard Gaussians are smooth low-pass primitives, modeling edges and fine textures requires multiple overlapping, well-aligned splats, which creates severe bottlenecks during rasterization. To address this, we introduce Resonant Brane Splatting (RBS), a feed-forward ASR framework. RBS replaces flat Gaussians with Branes: expressive primitives that emit spatially varying colors to natively model local contrast and complex textures within a single footprint. We achieve this by augmenting the standard Gaussian envelope with internal Gaussian-Hermite modes, assigning a distinct color coefficient to each. The zero-order mode recovers standard GS, while higher-order modes capture high frequencies. We predict Brane parameters directly from low-resolution features. Because Branes provide a mathematically richer formulation than simple Gaussians, far fewer primitives need to overlap to reconstruct a given target pixel. To exploit this, we introduce an efficient fully differentiable rasterizer with a precise culling strategy based on the classical quantum turning point. This allows us to safely skip negligible regions, drastically reducing the rendering overhead. Experiments on standard ASR benchmarks show that RBS improves reconstruction quality over implicit and GS baselines, while achieving superior speed-quality trade-off than prior GS methods.

Problem

• Arbitrary-Scale Super-Resolution
• Gaussian Splatting
• image reconstruction
• edge modeling
• texture synthesis

Innovation

• Resonant Brane Splatting
• Arbitrary-Scale Super-Resolution
• Gaussian-Hermite modes
• differentiable rasterization
• expressive primitives