DANICE: Domain adaptation without forgetting in neural image compression

TL;DR

This paper introduces DANICE, a neural image compression method that adapts to new domains without losing original capabilities, by adding minimal parameters to prevent forgetting and maintain compatibility.

Contribution

The paper proposes CAwF, a novel framework for domain adaptation in neural image compression that avoids catastrophic forgetting by preserving the source codec during adaptation.

Findings

Effective domain adaptation with minimal data

Prevents forgetting of original coding capabilities

Maintains compatibility with previous bitstreams

Abstract

Neural image compression (NIC) is a new coding paradigm where coding capabilities are captured by deep models learned from data. This data-driven nature enables new potential functionalities. In this paper, we study the adaptability of codecs to custom domains of interest. We show that NIC codecs are transferable and that they can be adapted with relatively few target domain images. However, naive adaptation interferes with the solution optimized for the original source domain, resulting in forgetting the original coding capabilities in that domain, and may even break the compatibility with previously encoded bitstreams. Addressing these problems, we propose Codec Adaptation without Forgetting (CAwF), a framework that can avoid these problems by adding a small amount of custom parameters, where the source codec remains embedded and unchanged during the adaptation process. Experiments demonstrate its effectiveness and provide useful insights on the characteristics of catastrophic interference in NIC.