Showing 1–1 of 1 results for author: Glazov, F

Accelerated Deep Lossless Image Coding with Unified Paralleleized GPU Coding Architecture

Authors: Benjamin Lukas Cajus Barzen, Fedor Glazov, Jonas Geistert, Thomas Sikora

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… ▽ More 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. △ Less

Submitted 11 July, 2022; originally announced July 2022.