SAPNet: Segmentation-Aware Progressive Network for Perceptual Contrastive Deraining

TL;DR

SAPNet is a novel segmentation-aware progressive network that leverages contrastive learning to improve single image deraining, effectively preserving semantic details and enhancing performance on synthetic and real-world datasets.

Contribution

The paper introduces a segmentation-aware progressive network with contrastive learning and an unsupervised background segmentation module for improved deraining and semantic preservation.

Findings

Outperforms top existing deraining methods on benchmarks.

Enhances object detection and semantic segmentation post-deraining.

Uses contrastive loss to better align derained images with groundtruth in semantic space.

Abstract

Deep learning algorithms have recently achieved promising deraining performances on both the natural and synthetic rainy datasets. As an essential low-level pre-processing stage, a deraining network should clear the rain streaks and preserve the fine semantic details. However, most existing methods only consider low-level image restoration. That limits their performances at high-level tasks requiring precise semantic information. To address this issue, in this paper, we present a segmentation-aware progressive network (SAPNet) based upon contrastive learning for single image deraining. We start our method with a lightweight derain network formed with progressive dilated units (PDU). The PDU can significantly expand the receptive field and characterize multi-scale rain streaks without the heavy computation on multi-scale images. A fundamental aspect of this work is an unsupervised background segmentation (UBS) network initialized with ImageNet and Gaussian weights. The UBS can faithfully preserve an image's semantic information and improve the generalization ability to unseen photos. Furthermore, we introduce a perceptual contrastive loss (PCL) and a learned perceptual image similarity loss (LPISL) to regulate model learning. By exploiting the rainy image and groundtruth as the negative and the positive sample in the VGG-16 latent space, we bridge the fine semantic details between the derained image and the groundtruth in a fully constrained manner. Comprehensive experiments on synthetic and real-world rainy images show our model surpasses top-performing methods and aids object detection and semantic segmentation with considerable efficacy. A Pytorch Implementation is available at https://github.com/ShenZheng2000/SAPNet-for-image-deraining.