Iterative Gradient Encoding Network with Feature Co-Occurrence Loss for Single Image Reflection Removal

TL;DR

This paper introduces an iterative gradient encoding network with a novel feature co-occurrence loss to effectively remove reflections from single images, demonstrating superior performance across diverse reflection types and strengths.

Contribution

The study presents a new iterative network architecture and a feature co-occurrence loss that improve generalization and reflection removal quality in single image reflection removal tasks.

Findings

Outperforms state-of-the-art methods on SIR$^2$ dataset

Effectively removes reflections across diverse backgrounds

Maintains performance even with strong reflections

Abstract

Removing undesired reflections from a photo taken in front of glass is of great importance for enhancing visual computing systems' efficiency. Previous learning-based approaches have produced visually plausible results for some reflections type, however, failed to generalize against other reflection types. There is a dearth of literature for efficient methods concerning single image reflection removal, which can generalize well in large-scale reflection types. In this study, we proposed an iterative gradient encoding network for single image reflection removal. Next, to further supervise the network in learning the correlation between the transmission layer features, we proposed a feature co-occurrence loss. Extensive experiments on the public benchmark dataset of SIR$^2$ demonstrated that our method can remove reflection favorably against the existing state-of-the-art method on all imaging settings, including diverse backgrounds. Moreover, as the reflection strength increases, our method can still remove reflection even where other state of the art methods failed.