Indirect Domain Shift for Single Image Dehazing

TL;DR

This paper introduces an indirect domain shift approach with explicit constraints in CNNs to improve single image dehazing, effectively recovering fine textures by narrowing the domain gap between hazy and clear images.

Contribution

It proposes a novel indirect domain shift mechanism with explicit constraints and two training schemes, enhancing CNN dehazing performance over existing methods.

Findings

Outperforms state-of-the-art dehazing methods

Effective in recovering fine texture details

Validates the proposed approach through extensive experiments

Abstract

Despite their remarkable expressibility, convolution neural networks (CNNs) still fall short of delivering satisfactory results on single image dehazing, especially in terms of faithful recovery of fine texture details. In this paper, we argue that the inadequacy of conventional CNN-based dehazing methods can be attributed to the fact that the domain of hazy images is too far away from that of clear images, rendering it difficult to train a CNN for learning direct domain shift through an end-to-end manner and recovering texture details simultaneously. To address this issue, we propose to add explicit constraints inside a deep CNN model to guide the restoration process. In contrast to direct learning, the proposed mechanism shifts and narrows the candidate region for the estimation output via multiple confident neighborhoods. Therefore, it is capable of consolidating the expressibility of different architectures, resulting in a more accurate indirect domain shift (IDS) from the hazy images to that of clear images. We also propose two different training schemes, including hard IDS and soft IDS, which further reveal the effectiveness of the proposed method. Our extensive experimental results indicate that the dehazing method based on this mechanism outperforms the state-of-the-arts.