Learned Lossless Image Compression Through Interpolation With Low Complexity

TL;DR

This paper introduces a scale-based auto-regressive image compression method using interpolation that achieves competitive performance with significantly reduced neural network complexity and faster processing times.

Contribution

It proposes an interpolation-based lossless image compression approach with shared neural networks and optimized architecture, reducing complexity while maintaining high compression quality.

Findings

Achieves comparable or better compression than recent models.

Requires over 10x fewer neural network parameters.

Offers significantly shorter encoding and decoding times.

Abstract

With the increasing popularity of deep learning in image processing, many learned lossless image compression methods have been proposed recently. One group of algorithms that have shown good performance are based on learned pixel-based auto-regressive models, however, their sequential nature prevents easily parallelized computations and leads to long decoding times. Another popular group of algorithms are based on scale-based auto-regressive models and can provide competitive compression performance while also enabling simple parallelization and much shorter decoding times. However, their major drawback are the used large neural networks and high computational complexity. This paper presents an interpolation based learned lossless image compression method which falls in the scale-based auto-regressive models group. The method achieves better than or on par compression performance with the recent scale-based auto-regressive models, yet requires more than 10x less neural network parameters and encoding/decoding computation complexity. These achievements are due to the contributions/findings in the overall system and neural network architecture design, such as sharing interpolator neural networks across different scales, using separate neural networks for different parameters of the probability distribution model and performing the processing in the YCoCg-R color space instead of the RGB color space.