提出可逆解耦网络，有效分离图像中透射与反射成分。
Reversible Decoupling Network for Single Image Reflection Removal
- 通过可逆编码器保留关键信息，动态解耦透射与反射特征。
- 在五个基准数据集上超越当前最优方法，NTIRE 2025 挑战赛双项第一。
- 适合图像去反射任务研究者及实际应用开发者参考。
基于深度学习的单图反射去除方法近年来取得显著进展，主要归因于两点：1）利用预训练识别特征作为输入；2）设计双流交互网络。然而，根据信息瓶颈原理，高层语义线索在逐层传播中容易被压缩或丢弃。此外，双流网络中的交互模式在各层间固定，限制了整体性能。为此，本文提出一种新型架构——可逆解耦网络（RDNet），其在前向传播中采用可逆编码器，确保关键信息不丢失，并灵活解耦透射与反射相关特征。同时，设计传输率感知提示生成器，动态校准特征，进一步提升性能。大量实验表明，RDNet 在五个广泛使用的基准数据集上优于现有最先进方法。在 NTIRE 2025 单图反射去除野外挑战赛中，无论保真度还是感知质量均排名第一。代码已开源。
原文摘要 · Abstract (English)
Recent deep-learning-based approaches to single-image reflection removal have shown promising advances, primarily for two reasons: 1) the utilization of recognition-pretrained features as inputs, and 2) the design of dual-stream interaction networks. However, according to the Information Bottleneck principle, high-level semantic clues tend to be compressed or discarded during layer-by-layer propagation. Additionally, interactions in dual-stream networks follow a fixed pattern across different layers, limiting overall performance. To address these limitations, we propose a novel architecture called Reversible Decoupling Network (RDNet), which employs a reversible encoder to secure valuable information while flexibly decoupling transmission- and reflection-relevant features during the forward pass. Furthermore, we customize a transmission-rate-aware prompt generator to dynamically calibrate features, further boosting performance. Extensive experiments demonstrate the superiority of RDNet over existing SOTA methods on five widely-adopted benchmark datasets. RDNet achieves the best performance in the NTIRE 2025 Single Image Reflection Removal in the Wild Challenge in both fidelity and perceptual comparison. Our code is available at https://github.com/lime-j/RDNet
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。