Single Image Reflection Removal Using Deep Encoder-Decoder Network

TL;DR

This paper introduces a deep encoder-decoder neural network trained on synthetic data to effectively remove reflections from single images, outperforming existing methods on real-world images.

Contribution

It presents a novel deep learning approach for reflection removal that learns from synthetic data and generalizes well to real-world images.

Findings

Outperforms state-of-the-art techniques on real-world images

Effective despite training only on synthetic data

Significantly improves reflection removal quality

Abstract

Image of a scene captured through a piece of transparent and reflective material, such as glass, is often spoiled by a superimposed layer of reflection image. While separating the reflection from a familiar object in an image is mentally not difficult for humans, it is a challenging, ill-posed problem in computer vision. In this paper, we propose a novel deep convolutional encoder-decoder method to remove the objectionable reflection by learning a map between image pairs with and without reflection. For training the neural network, we model the physical formation of reflections in images and synthesize a large number of photo-realistic reflection-tainted images from reflection-free images collected online. Extensive experimental results show that, although the neural network learns only from synthetic data, the proposed method is effective on real-world images, and it significantly outperforms the other tested state-of-the-art techniques.