Near-lossless $\ell_\infty$-constrained Image Decompression via Deep Neural Network

TL;DR

This paper introduces a deep neural network approach for near-lossless image decompression that emphasizes $lu_infty$ fidelity, effectively restoring subtle details often lost in traditional methods, with applications in sensitive fields.

Contribution

The paper proposes a novel neural network design incorporating an $lu_infty$ fidelity criterion, improving the preservation of distinctive image details during decompression.

Findings

Outperforms state-of-the-art methods in $lu_infty$ error and perceptual quality

Restores subtle image details often missed by other algorithms

Competitive in $lu_infty$ and $lu_2$ error metrics

Abstract

Recently a number of CNN-based techniques were proposed to remove image compression artifacts. As in other restoration applications, these techniques all learn a mapping from decompressed patches to the original counterparts under the ubiquitous $\ell_\infty$ metric. However, this approach is incapable of restoring distinctive image details which may be statistical outliers but have high semantic importance (e.g., tiny lesions in medical images). To overcome this weakness, we propose to incorporate an $\ell_\infty$ fidelity criterion in the design of neural network so that no small, distinctive structures of the original image can be dropped or distorted. Experimental results demonstrate that the proposed method outperforms the state-of-the-art methods in $\ell_\infty$ error metric and perceptual quality, while being competitive in $\ell_2$ error metric as well. It can restore subtle image details that are otherwise destroyed or missed by other algorithms. Our research suggests a new machine learning paradigm of ultra high fidelity image compression that is ideally suited for applications in medicine, space, and sciences.