Learned Lossless Image Compression With Combined Autoregressive Models And Attention Modules

TL;DR

This paper introduces a novel lossless image compression method that combines autoregressive models and attention modules, inspired by lossy compression techniques, achieving superior performance over classical and existing learning-based methods.

Contribution

The paper proposes a new lossless compression architecture integrating GMM, attention modules, and autoregressive models, inspired by lossy compression strategies.

Findings

Outperforms most classical lossless methods

Surpasses existing learning-based approaches

Demonstrates the effectiveness of combining GMM with attention and autoregressive models

Abstract

Lossless image compression is an essential research field in image compression. Recently, learning-based image compression methods achieved impressive performance compared with traditional lossless methods, such as WebP, JPEG2000, and FLIF. However, there are still many impressive lossy compression methods that can be applied to lossless compression. Therefore, in this paper, we explore the methods widely used in lossy compression and apply them to lossless compression. Inspired by the impressive performance of the Gaussian mixture model (GMM) shown in lossy compression, we generate a lossless network architecture with GMM. Besides noticing the successful achievements of attention modules and autoregressive models, we propose to utilize attention modules and add an extra autoregressive model for raw images in our network architecture to boost the performance. Experimental results show that our approach outperforms most classical lossless compression methods and existing learning-based methods.