Reflection Separation Using Guided Annotation
This work addresses the problem of reflection separation for photographers and computer vision applications, offering a more viable annotation process for textured images compared to existing sparsity-based methods, though it is incremental in improving user interaction.
The paper tackled the ill-posed problem of separating reflection and transmission layers in images taken through glass by introducing a method that uses a Gaussian Mixture Model patch prior to identify likely decomposition modes, reducing user annotation to selecting from sparse patches, and demonstrated performance on synthesized and real images.
Photographs taken through a glass surface often contain an approximately linear superposition of reflected and transmitted layers. Decomposing an image into these layers is generally an ill-posed task and the use of an additional image prior and user provided cues is presently necessary in order to obtain good results. Current annotation approaches rely on a strong sparsity assumption. For images with significant texture this assumption does not typically hold, thus rendering the annotation process unviable. In this paper we show that using a Gaussian Mixture Model patch prior, the correct local decomposition can almost always be found as one of 100 likely modes of the posterior. Thus, the user need only choose one of these modes in a sparse set of patches and the decomposition may then be completed automatically. We demonstrate the performance of our method using synthesized and real reflection images.