Learning to Inpaint for Image Compression

TL;DR

This paper introduces architectural improvements for deep lossy image compression, demonstrating that multi-stage residual prediction and inpainting significantly reduce file sizes while maintaining quality.

Contribution

The paper proposes two novel architectural strategies—residual prediction and inpainting—improving compression efficiency in deep image encoders.

Findings

Over 60% reduction in file size with similar quality

Residual prediction enhances learning and content approximation

Inpainting reduces information needed for high-quality reconstruction

Abstract

We study the design of deep architectures for lossy image compression. We present two architectural recipes in the context of multi-stage progressive encoders and empirically demonstrate their importance on compression performance. Specifically, we show that: (a) predicting the original image data from residuals in a multi-stage progressive architecture facilitates learning and leads to improved performance at approximating the original content and (b) learning to inpaint (from neighboring image pixels) before performing compression reduces the amount of information that must be stored to achieve a high-quality approximation. Incorporating these design choices in a baseline progressive encoder yields an average reduction of over $60\%$ in file size with similar quality compared to the original residual encoder.