Learning Arbitrary-Scale RAW Image Downscaling with Wavelet-based Recurrent Reconstruction

TL;DR

This paper introduces a wavelet-based recurrent framework for arbitrary-scale RAW image downscaling that preserves details and textures, outperforming existing methods, and provides a new dataset for benchmarking.

Contribution

The paper proposes a novel wavelet-based recurrent reconstruction method for arbitrary-scale RAW image downscaling, including new modules and a dataset for non-integer scaling.

Findings

Outperforms state-of-the-art methods quantitatively.

Achieves superior visual quality in reconstructed images.

Introduces a new non-integer downscaling dataset for benchmarking.

Abstract

Image downscaling is critical for efficient storage and transmission of high-resolution (HR) images. Existing learning-based methods focus on performing downscaling within the sRGB domain, which typically suffers from blurred details and unexpected artifacts. RAW images, with their unprocessed photonic information, offer greater flexibility but lack specialized downscaling frameworks. In this paper, we propose a wavelet-based recurrent reconstruction framework that leverages the information lossless attribute of wavelet transformation to fulfill the arbitrary-scale RAW image downscaling in a coarse-to-fine manner, in which the Low-Frequency Arbitrary-Scale Downscaling Module (LASDM) and the High-Frequency Prediction Module (HFPM) are proposed to preserve structural and textural integrity of the reconstructed low-resolution (LR) RAW images, alongside an energy-maximization loss to align high-frequency energy between HR and LR domain. Furthermore, we introduce the Realistic Non-Integer RAW Downscaling (Real-NIRD) dataset, featuring a non-integer downscaling factor of 1.3$\times$, and incorporate it with publicly available datasets with integer factors (2$\times$, 3$\times$, 4$\times$) for comprehensive benchmarking arbitrary-scale image downscaling purposes. Extensive experiments demonstrate that our method outperforms existing state-of-the-art competitors both quantitatively and visually. The code and dataset will be released at https://github.com/RenYangSCU/ASRD.