Semantic Guided Single Image Reflection Removal

Teaser Image


Reflection is common in images capturing scenes behind a glass window, which is not only a disturbance visually but also influence the performance of other computer vision al- gorithms. Single image reflection removal is an ill-posed problem because the color at each pixel needs to be sepa- rated into two values, i.e., the desired clear background and the reflection. To solve it, existing methods propose priors such as smoothness, color consistency. However, the low- level priors are not reliable in complex scenes, for instance, when capturing a real outdoor scene through a window, both the foreground and background contain both smooth and sharp area and a variety of color. In this paper, in- spired by the fact that human can separate the two layers easily by recognizing the objects, we use the object seman- tic as guidance to force the same semantic object belong to the same layer. Extensive experiments on different datasets show that adding the semantic information offers a signifi- cant improvement to reflection separation. We also demon- strate the applications of the proposed method to other com- puter vision tasks

The ACM Transactions on Multimedia Computing, Communications, and Applications