Cycle-Consistent Tuning for Layered Image Decomposition

TL;DR

This paper introduces a cycle-consistent tuning framework using diffusion models for layered image decomposition, effectively disentangling complex visual layers like logos from surfaces with improved robustness and generalization.

Contribution

It proposes a novel cycle-consistent fine-tuning method with progressive self-improvement for layered image decomposition using diffusion models, addressing complex interactions.

Findings

Achieves accurate and coherent layer separation.

Generalizes well across different decomposition tasks.

Enhances robustness through cycle consistency and self-improvement.

Abstract

Disentangling visual layers in real-world images is a persistent challenge in vision and graphics, as such layers often involve non-linear and globally coupled interactions, including shading, reflection, and perspective distortion. In this work, we present an in-context image decomposition framework that leverages large diffusion foundation models for layered separation. We focus on the challenging case of logo-object decomposition, where the goal is to disentangle a logo from the surface on which it appears while faithfully preserving both layers. Our method fine-tunes a pretrained diffusion model via lightweight LoRA adaptation and introduces a cycle-consistent tuning strategy that jointly trains decomposition and composition models, enforcing reconstruction consistency between decomposed and recomposed images. This bidirectional supervision substantially enhances robustness in cases where the layers exhibit complex interactions. Furthermore, we introduce a progressive self-improving process, which iteratively augments the training set with high-quality model-generated examples to refine performance. Extensive experiments demonstrate that our approach achieves accurate and coherent decompositions and also generalizes effectively across other decomposition types, suggesting its potential as a unified framework for layered image decomposition.