Abstract:Pansharpening refers to fusing a low-resolution multispectral image (LRMS) and a high-resolution panchromatic (PAN) image to generate a high-resolution multispectral image (HRMS). Traditional pansharpening methods use a single pair of LRMS and PAN to generate HRMS at full resolution, but they fail to generate high-quality fused products due to the assumption of a (often inaccurate) linear relationship between the fused products. Convolutional neural network methods, i.e. , supervised and unsupervised learning approaches, can model any arbitrary non-linear relationship among data, but performing even worse than traditional methods when testing data are not consistent with training data. Moreover, supervised methods rely on simulating reduced resolution data for training causing information loss at full resolution. Unsupervised pansharpening suffers from distortion due to the lack of reference images and inaccuracy in the estimation of the degradation process. In this paper, we propose a zero-shot semi-supervised method for pansharpening (ZS-Pan), which only requires a single pair of PAN/LRMS images for training and testing networks combining both the pros of supervised and unsupervised methods. Facing with challenges of limited training data and no reference images, the ZS-Pan framework is built with a two-phase three-component model, i.e. , the reduced resolution supervised pre-training (RSP), the spatial degradation establishment (SDE), and the full resolution unsupervised generation (FUG) stages. Specifically, a special parameter initialization technique, a data augmentation strategy, and a non-linear degradation network are proposed to improve the representation ability of the network. In our experiments, we evaluate the performance of the proposed framework on different datasets using some state-of-the-art (SOTA) pansharpening approaches for comparison. Results show that our ZS-Pan outperforms these SOTA methods, both visually and quantitatively. Codes will be available after possible acceptance.

Training Pansharpening Networks at Full Resolution Using Degenerate Invariance

Efficient and Accurate Hyperspectral Pansharpening Using 3D VolumeNet and 2.5D Texture Transfer

Zero-shot semi-supervised learning for pansharpening

An Unsupervised CNN-Based Pansharpening Framework with Spectral-Spatial Fidelity Balance

A Multi-Stage Progressive Pansharpening Network Based on Detail Injection with Redundancy Reduction

UPanGAN: Unsupervised pansharpening based on the spectral and spatial loss constrained Generative Adversarial Network

Pan-Sharpening with Customized Transformer and Invertible Neural Network

Unsupervised Deep Learning-based Pansharpening with Jointly-Enhanced Spectral and Spatial Fidelity

Pannet: A Deep Network Architecture For Pan-Sharpening

Unsupervised Pansharpening Based on Double-Cycle Consistency

Effective Pan-Sharpening with Transformer and Invertible Neural Network

Detail Information Prior Net for Remote Sensing Image Pansharpening

Pansharpening by Combining Enhanced Spatial-Spectral Fidelity and Gradient-Domain Guided Filtering

An Unsupervised Generative Adversarial Network Based on U-Shaped Structure for Pansharpening

Band-wise Hyperspectral Image Pansharpening using CNN Model Propagation

Deep Gradient Projection Networks for Pan-sharpening

A New Pansharpening Method Using Objectness Based Saliency Analysis and Saliency Guided Deep Residual Network

Pan-Sharpening Based On Parallel Pyramid Convolutional Neural Network

Pan-sharpening Method Based on Unsupervised and Divide-and-conquer Strategies

Convolutional Neural Network for Pansharpening with Spatial Structure Enhancement Operator

Towards Robust Pansharpening: A Large-Scale High-Resolution Multi-Scene Dataset and Novel Approach