Low-complexity Overfitted Neural Image Codec

TL;DR

This paper introduces a low-complexity neural image codec that overfits decoder parameters to individual images, achieving competitive performance with significantly fewer computations than traditional autoencoders and surpassing HEVC in various conditions.

Contribution

It presents a novel overfitted neural image codec with drastically reduced complexity, matching autoencoder performance and improving rate-distortion efficiency.

Findings

Achieves similar quality to autoencoders with 2300 multiplications per pixel

Surpasses HEVC performance under various coding conditions

Reduces rate by 14% compared to previous overfitted codecs

Abstract

We propose a neural image codec at reduced complexity which overfits the decoder parameters to each input image. While autoencoders perform up to a million multiplications per decoded pixel, the proposed approach only requires 2300 multiplications per pixel. Albeit low-complexity, the method rivals autoencoder performance and surpasses HEVC performance under various coding conditions. Additional lightweight modules and an improved training process provide a 14% rate reduction with respect to previous overfitted codecs, while offering a similar complexity. This work is made open-source at https://orange-opensource.github.io/Cool-Chic/