IICNet: A Generic Framework for Reversible Image Conversion

TL;DR

IICNet is a versatile invertible neural network framework that enables reversible image conversion, effectively preserving information and outperforming task-specific methods across various visual content transformation tasks.

Contribution

The paper introduces IICNet, a generic invertible neural network framework for reversible image conversion, eliminating the need for task-specific network design.

Findings

Outperforms existing RIC methods on multiple tasks

Demonstrates strong generalization to new RIC tasks

Maintains high information preservation during conversion

Abstract

Reversible image conversion (RIC) aims to build a reversible transformation between specific visual content (e.g., short videos) and an embedding image, where the original content can be restored from the embedding when necessary. This work develops Invertible Image Conversion Net (IICNet) as a generic solution to various RIC tasks due to its strong capacity and task-independent design. Unlike previous encoder-decoder based methods, IICNet maintains a highly invertible structure based on invertible neural networks (INNs) to better preserve the information during conversion. We use a relation module and a channel squeeze layer to improve the INN nonlinearity to extract cross-image relations and the network flexibility, respectively. Experimental results demonstrate that IICNet outperforms the specifically-designed methods on existing RIC tasks and can generalize well to various newly-explored tasks. With our generic IICNet, we no longer need to hand-engineer task-specific embedding networks for rapidly occurring visual content. Our source codes are available at: https://github.com/felixcheng97/IICNet.