Hierarchy Flow For High-Fidelity Image-to-Image Translation

TL;DR

Hierarchy Flow is a novel flow-based model that significantly improves content preservation in high-fidelity image-to-image translation tasks, outperforming existing methods across various benchmarks.

Contribution

The paper introduces Hierarchy Flow, a new hierarchical coupling design and aligned-style loss to enhance content preservation in flow-based I2I translation models.

Findings

Achieves state-of-the-art results on multiple I2I benchmarks.

Excels in both strong- and normal-fidelity translation tasks.

Demonstrates better content preservation compared to existing methods.

Abstract

Image-to-image (I2I) translation comprises a wide spectrum of tasks. Here we divide this problem into three levels: strong-fidelity translation, normal-fidelity translation, and weak-fidelity translation, indicating the extent to which the content of the original image is preserved. Although existing methods achieve good performance in weak-fidelity translation, they fail to fully preserve the content in both strong- and normal-fidelity tasks, e.g. sim2real, style transfer and low-level vision. In this work, we propose Hierarchy Flow, a novel flow-based model to achieve better content preservation during translation. Specifically, 1) we first unveil the drawbacks of standard flow-based models when applied to I2I translation. 2) Next, we propose a new design, namely hierarchical coupling for reversible feature transformation and multi-scale modeling, to constitute Hierarchy Flow. 3) Finally, we present a dedicated aligned-style loss for a better trade-off between content preservation and stylization during translation. Extensive experiments on a wide range of I2I translation benchmarks demonstrate that our approach achieves state-of-the-art performance, with convincing advantages in both strong- and normal-fidelity tasks. Code and models will be at https://github.com/WeichenFan/HierarchyFlow.