UCL-Dehaze: Towards Real-world Image Dehazing via Unsupervised Contrastive Learning

TL;DR

UCL-Dehaze introduces an unsupervised contrastive learning approach for real-world image dehazing that leverages unpaired data and adversarial training to improve performance without requiring paired datasets.

Contribution

It proposes a novel unsupervised contrastive learning framework with a self-contrastive perceptual loss for effective real-world image dehazing without paired training data.

Findings

Outperforms recent dehazing methods on benchmark tests.

Achieves superior results using only 1,800 unpaired images.

Effectively bridges the domain gap between synthetic and real-world haze.

Abstract

While the wisdom of training an image dehazing model on synthetic hazy data can alleviate the difficulty of collecting real-world hazy/clean image pairs, it brings the well-known domain shift problem. From a different yet new perspective, this paper explores contrastive learning with an adversarial training effort to leverage unpaired real-world hazy and clean images, thus bridging the gap between synthetic and real-world haze is avoided. We propose an effective unsupervised contrastive learning paradigm for image dehazing, dubbed UCL-Dehaze. Unpaired real-world clean and hazy images are easily captured, and will serve as the important positive and negative samples respectively when training our UCL-Dehaze network. To train the network more effectively, we formulate a new self-contrastive perceptual loss function, which encourages the restored images to approach the positive samples and keep away from the negative samples in the embedding space. Besides the overall network architecture of UCL-Dehaze, adversarial training is utilized to align the distributions between the positive samples and the dehazed images. Compared with recent image dehazing works, UCL-Dehaze does not require paired data during training and utilizes unpaired positive/negative data to better enhance the dehazing performance. We conduct comprehensive experiments to evaluate our UCL-Dehaze and demonstrate its superiority over the state-of-the-arts, even only 1,800 unpaired real-world images are used to train our network. Source code has been available at https://github.com/yz-wang/UCL-Dehaze.