COMPASS: High-Efficiency Deep Image Compression with Arbitrary-scale Spatial Scalability

TL;DR

COMPASS is a neural network-based image compression method that enables arbitrary-scale spatial scalability with high efficiency, outperforming existing scalable and non-scalable methods in terms of compression rate.

Contribution

This paper introduces COMPASS, a novel neural network framework supporting flexible, arbitrary-scale spatially scalable image compression with an innovative inter-layer scale prediction method.

Findings

Achieves BD-rate gains of -58.33% and -47.17% compared to SHVC and other NN methods.

Supports arbitrary scale factors with comparable or better efficiency than single-layer coding.

Demonstrates superior performance across various scale factor combinations.

Abstract

Recently, neural network (NN)-based image compression studies have actively been made and has shown impressive performance in comparison to traditional methods. However, most of the works have focused on non-scalable image compression (single-layer coding) while spatially scalable image compression has drawn less attention although it has many applications. In this paper, we propose a novel NN-based spatially scalable image compression method, called COMPASS, which supports arbitrary-scale spatial scalability. Our proposed COMPASS has a very flexible structure where the number of layers and their respective scale factors can be arbitrarily determined during inference. To reduce the spatial redundancy between adjacent layers for arbitrary scale factors, our COMPASS adopts an inter-layer arbitrary scale prediction method, called LIFF, based on implicit neural representation. We propose a combined RD loss function to effectively train multiple layers. Experimental results show that our COMPASS achieves BD-rate gain of -58.33% and -47.17% at maximum compared to SHVC and the state-of-the-art NN-based spatially scalable image compression method, respectively, for various combinations of scale factors. Our COMPASS also shows comparable or even better coding efficiency than the single-layer coding for various scale factors.