Upsampling Improvement for Overfitted Neural Coding

TL;DR

This paper introduces a new upsampling method for neural image codecs that reduces complexity and parameters, leading to a 4.7% rate reduction in overfitted auto-encoder based compression.

Contribution

A novel upsampling structure is proposed to improve neural coding efficiency by reducing complexity and parameters in overfitted auto-encoder frameworks.

Findings

Achieved 4.7% rate reduction with the new upsampling method.

Demonstrated improved performance within the Cool-Chic coding framework.

Provided open-source code for the proposed method.

Abstract

Neural image compression, based on auto-encoders and overfitted representations, relies on a latent representation of the coded signal. This representation needs to be compact and uses low resolution feature maps. In the decoding process, those latents are upsampled and filtered using stacks of convolution filters and non linear elements to recover the decoded image. Therefore, the upsampling process is crucial in the design of a neural coding scheme and is of particular importance for overfitted codecs where the network parameters, including the upsampling filters, are part of the representation. This paper addresses the improvement of the upsampling process in order to reduce its complexity and limit the number of parameters. A new upsampling structure is presented whose improvements are illustrated within the Cool-Chic overfitted image coding framework. The proposed approach offers a rate reduction of 4.7%. The code is provided.