ReLo-IRR: Reflection-Guided LoRA Framework for Image Reflection Removal
2026-07-03 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors present ReLo-IRR, a new method to remove reflections from single images more effectively. Unlike previous approaches that treat all reflections the same, their technique first measures how strong the reflection is, then adjusts the removal process accordingly. This helps avoid leftover reflections in strong cases and detail loss in weak ones. Their approach also links this reflection strength to how the image is cleaned over time, improving overall results. Tests show ReLo-IRR works well across various challenging reflection conditions.
Single-image reflection removalDiffusion modelsLoRA (Low-Rank Adaptation)Reflection strength estimationRectified flow modelTimestep embeddingsDenoising processControlNetImage enhancement
Authors
Chaoqun Wang, Yuehuan Wei, Haoxiang Cao, Shaobo Min
Abstract
Single-image reflection removal (SIRR) aims to recover the clean transmission layer from a reflection-contaminated image. Although recent methods achieve promising results with large diffusion models, they rely on image-agnostic adaptation strategies, e.g., fine-tuning or ControlNet, that enforce uniform suppression regardless of reflection severity. As a result, heavy reflections often leave residuals, while weak ones suffer from detail loss. To this end, we propose ReLo-IRR, a reflection-guided LoRA framework built upon the rectified flow model. First, a lightweight estimator is designed to predict the reflection strength descriptor, providing an explicit prior of reflection dominance for each image and enabling image-dependent LoRA modulation. Second, we introduce a time-conditioned mechanism that fuses this reflection descriptor with timestep embeddings, enabling LoRA modulation to evolve consistently with the coarse-to-fine denoising process. By jointly modeling reflection strength and denoising dynamics, our ReLo-IRR achieves robust suppression of diverse reflection conditions. Extensive experiments on challenging benchmarks validate the effectiveness of ReLo-IRR, demonstrating superior dereflection performance and robust generalization. The code is released at https://github.com/KONGBAI-8080/ReLo-IRR.