Convolutional Autoencoders for Lossy Light Field Compression

TL;DR

This paper presents a convolutional autoencoder-based method for lossy compression of 4D Light Field data, achieving high compression ratios with minimal quality loss and real-time encoding and decoding capabilities.

Contribution

It introduces an efficient neural network architecture specifically designed for 4D Light Field compression, achieving a 48.6x compression ratio with high quality metrics.

Findings

Achieved 48.6x compression with PSNR of 29.46 dB

Real-time encoding and decoding speeds of around 1.6-1.8 seconds

Network size is 584MB, suitable for practical use

Abstract

Expansion and reduction of a neural network's width has well known properties in terms of the entropy of the propagating information. When carefully stacked on top of one another, an encoder network and a decoder network produce an autoencoder, often used in compression. Using this architecture, we develop an efficient method of encoding and decoding 4D Light Field data, with a substantial compression factor at a minimal loss in quality. Our best results managed to achieve a compression of 48.6x, with a PSNR of 29.46 dB and a SSIM of 0.8104. Computations of the encoder and decoder can be run in real time, with average computation times of 1.62s and 1.81s respectively, and the entire network occupies a reasonable 584MB by today's storage standards.