The paper "Progressive High-Frequency Reconstruction for Pan-Sharpening with Implicit Neural Representation" addresses the challenge of enhancing the resolution of multispectral (MS) images by leveraging the high-frequency signals from panchromatic (PAN) images. Traditional deep neural networks (DNNs) often prioritize low-frequency components during training, leading to limited restoration of high-frequency edge details in MS images. To overcome this, the authors propose a novel method that treats pan-sharpening as a coarse-to-fine high-frequency restoration problem. They introduce a Band-limited Multi-scale High-frequency Generator (BMHG) to generate high-frequency signals from the PAN image within different bandwidths, which are progressively injected into the MS image during training. Additionally, they employ a Spatial-spectral Implicit Image Function (SIIF) to effectively fuse spatial and spectral features in the continuous domain, addressing pixel position misalignment issues. Extensive experiments on various datasets demonstrate that the proposed method outperforms existing approaches in terms of both quantitative and visual metrics. The key contributions include a novel pan-sharpening method, a multi-stage BMHG for generating multi-scale high-frequency signals, and an SIIF for continuous domain feature fusion.The paper "Progressive High-Frequency Reconstruction for Pan-Sharpening with Implicit Neural Representation" addresses the challenge of enhancing the resolution of multispectral (MS) images by leveraging the high-frequency signals from panchromatic (PAN) images. Traditional deep neural networks (DNNs) often prioritize low-frequency components during training, leading to limited restoration of high-frequency edge details in MS images. To overcome this, the authors propose a novel method that treats pan-sharpening as a coarse-to-fine high-frequency restoration problem. They introduce a Band-limited Multi-scale High-frequency Generator (BMHG) to generate high-frequency signals from the PAN image within different bandwidths, which are progressively injected into the MS image during training. Additionally, they employ a Spatial-spectral Implicit Image Function (SIIF) to effectively fuse spatial and spectral features in the continuous domain, addressing pixel position misalignment issues. Extensive experiments on various datasets demonstrate that the proposed method outperforms existing approaches in terms of both quantitative and visual metrics. The key contributions include a novel pan-sharpening method, a multi-stage BMHG for generating multi-scale high-frequency signals, and an SIIF for continuous domain feature fusion.