Accelerated Deep Lossless Image Coding with Unified Paralleleized GPU Coding Architecture

TL;DR

This paper introduces a GPU-accelerated neural network-based lossless image compression method that achieves fast encoding and decoding, outperforming existing algorithms like FLIF, and can be adapted to specific domains such as MRI images.

Contribution

The paper presents a novel parallelized GPU implementation of a neural network-based lossless image coder, enabling real-time performance and domain-specific training capabilities.

Findings

Encoding and decoding in less than one second on GPU

Outperforms FLIF in compression rate

Can be trained on domain-specific images like MRI

Abstract

We propose Deep Lossless Image Coding (DLIC), a full resolution learned lossless image compression algorithm. Our algorithm is based on a neural network combined with an entropy encoder. The neural network performs a density estimation on each pixel of the source image. The density estimation is then used to code the target pixel, beating FLIF in terms of compression rate. Similar approaches have been attempted. However, long run times make them unfeasible for real world applications. We introduce a parallelized GPU based implementation, allowing for encoding and decoding of grayscale, 8-bit images in less than one second. Because DLIC uses a neural network to estimate the probabilities used for the entropy coder, DLIC can be trained on domain specific image data. We demonstrate this capability by adapting and training DLIC with Magnet Resonance Imaging (MRI) images.