Semi-supervised Transfer Learning for Image Rain Removal

CVPR 2019  ·  Wei Wei, Deyu Meng, Qian Zhao, Zongben Xu, Ying Wu ·

Single image rain removal is a typical inverse problem in computer vision. The deep learning technique has been verified to be effective for this task and achieved state-of-the-art performance. However, previous deep learning methods need to pre-collect a large set of image pairs with/without synthesized rain for training, which tends to make the neural network be biased toward learning the specific patterns of the synthesized rain, while be less able to generalize to real test samples whose rain types differ from those in the training data. To this issue, this paper firstly proposes a semi-supervised learning paradigm toward this task. Different from traditional deep learning methods which only use supervised image pairs with/without synthesized rain, we further put real rainy images, without need of their clean ones, into the network training process. This is realized by elaborately formulating the residual between an input rainy image and its expected network output (clear image without rain) as a specific parametrized rain streaks distribution. The network is therefore trained to adapt real unsupervised diverse rain types through transferring from the supervised synthesized rain, and thus both the short-of-training-sample and bias-to-supervised-sample issues can be evidently alleviated. Experiments on synthetic and real data verify the superiority of our model compared to the state-of-the-arts.

PDF Abstract CVPR 2019 PDF CVPR 2019 Abstract

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Single Image Deraining Rain100H SEMI PSNR 16.56 # 13
SSIM 0.486 # 16
Single Image Deraining Rain100L SEMI PSNR 25.03 # 14
SSIM 0.842 # 18
Single Image Deraining Test100 SEMI PSNR 22.35 # 8
SSIM 0.788 # 12
Single Image Deraining Test1200 SEMI PSNR 26.05 # 9
SSIM 0.822 # 14
Single Image Deraining Test2800 SEMI PSNR 24.43 # 10
SSIM 0.782 # 11

Methods


No methods listed for this paper. Add relevant methods here