Abstract:Reflections often degrade the quality of the image by obstructing the background scene. This is not desirable for everyday users, and it negatively impacts the performance of multimedia applications that process images with reflections. Most current methods for removing reflections utilize supervised-learning models. However, these models require an extensive number of image pairs to perform well, especially on natural images with reflection, which is difficult to achieve in practice. In this paper, we propose a novel unsupervised framework for single-image reflection separation. Instead of learning from a large dataset, we optimize the parameters of two cross-coupled deep convolutional networks on a target image to generate two exclusive background and reflection layers. In particular, we design a new architecture of the network to embed semantic features extracted from a pre-trained deep classification network, which gives more meaningful separation similar to human perception. Quantitative and qualitative results on commonly used datasets in the literature show that our method's performance is at least on par with the state-of-the-art supervised methods and, occasionally, better without requiring large training datasets. Our results also show that our method significantly outperforms the closest unsupervised method in the literature for removing reflections from single images.

Separating Reflections From A Single Image Using Spatial Smoothness And Structure Information

Chromaticity-based separation of reflection components in a single image

Single Image Reflection Removal with Absorption Effect

Toward a General Model for Reflection Recovery and Single Image Enhancement.

Separation of Weak Reflection from a Single Superimposed Image Using Gradient Profile Sharpness

Separation of Weak Reflection from A Single Superimposed Image

PAR<SUP>2</SUP>Net: End-to-End Panoramic Image Reflection Removal

Simple and Efficient Method for Specularity Removal in an Image

Separation of superimposed images with unknown motions using sparsity priors

Reflection Scene Separation From A Single Image

Robust Separation of Reflection from Multiple Images

Automatic Reflection Removal Using Gradient Intensity and Motion Cues

Region-Aware Reflection Removal with Unified Content and Gradient Priors

Separation Of Diffuse And Specular Reflection In Color Images

Efficient Blind Separation Of Reflection Layers With Nonparametric Transformations

Reflection Separation Using Patch-Wise Sparse and Low-Rank Decomposition

A Robust Image Reflection Separation Method Based on Sift-Edge Flow.

Sparsity based reflection removal using external patch search

Unsupervised Single-Image Reflection Separation Using Perceptual Deep Image Priors

Single Image Reflection Removal With Physically-Based Training Images