Euler-inspired Decoupling Neural Operator for Efficient Pansharpening

TL;DR

This paper introduces EDNO, a physics-inspired neural operator for pansharpening that operates in the frequency domain, offering improved efficiency and spectral-spatial fidelity over existing deep learning methods.

Contribution

The paper proposes a novel Euler-inspired framework with explicit-implicit feature interaction modules, redefining pansharpening as a continuous frequency domain mapping.

Findings

EDNO achieves better efficiency-performance trade-offs than heavyweight architectures.

The frequency domain approach captures global features effectively.

The explicit-implicit interaction enhances spectral and spatial fidelity.

Abstract

Pansharpening aims to synthesize high-resolution multispectral (HR-MS) images by fusing the spatial textures of panchromatic (PAN) images with the spectral information of low-resolution multispectral (LR-MS) images. While recent deep learning paradigms, especially diffusion-based operators, have pushed the performance boundaries, they often encounter spectral-spatial blurring and prohibitive computational costs due to their stochastic nature and iterative sampling. In this paper, we propose the Euler-inspired Decoupling Neural Operator (EDNO), a physics-inspired framework that redefines pansharpening as a continuous functional mapping in the frequency domain. Departing from conventional Cartesian feature processing, our EDNO leverages Euler's formula to transform features into a polar coordinate system, enabling a novel explicit-implicit interaction mechanism. Specifically, we develop the Euler Feature Interaction Layer (EFIL), which decouples the fusion task into two specialized modules: 1) Explicit Feature Interaction Module, utilizing a linear weighting scheme to simulate phase rotation for adaptive geometric alignment; and 2) Implicit Feature Interaction Module, employing a feed-forward network to model spectral distributions for superior color consistency. By operating in the frequency domain, EDNO inherently captures global receptive fields while maintaining discretization-invariance. Experimental results on the three datasets demonstrate that EDNO offers a superior efficiency-performance balance compared to heavyweight architectures.