Details

Autor(en) / Beteiligte

• Wang, Guoqing
• Sun, Changming
• Sowmya, Arcot

Titel

Attentive Feature Refinement Network for Single Rainy Image Restoration

Ist Teil von

• IEEE transactions on image processing, 2021, Vol.30, p.3734-3747

Ort / Verlag

United States: IEEE

Erscheinungsjahr

2021

Link zum Volltext

• IEEE_Xplore

Quelle

IEEE Electronic Library Online

Beschreibungen/Notizen

• Despite the fact that great progress has been made on single image deraining tasks, it is still challenging for existing models to produce satisfactory results directly, and it often requires a single or multiple refinement stages to gradually improve the quality. However, in this paper, we demonstrate that existing image-level refinement with a stage-independent learning design is problematic with the side effect of over/under-deraining. To resolve this issue, we for the first time propose the mechanism of learning to carry out refinement on the unsatisfactory features, and propose a novel attentive feature refinement (AFR) module. Specifically, AFR is designed as a two-branched network for simultaneous rain-distribution-aware attention map learning and attention guided hierarchy-preserving feature refinement. Guided by task-specific attention, coarse features are progressively refined to better model the diversified rainy effects. By using a separable convolution as the basic component, our AFR module introduces little computation overhead and can be readily integrated into most rainy-to-clean image translation networks for achieving better deraining results. By incorporating a series of AFR modules into a general encoder-decoder network, AFR-Net is constructed for deraining and it achieves new state-of-the-art results on both synthetic and real images. Furthermore, by using AFR-Net as a teacher model, we explore the use of knowledge distillation to successfully learn a student model that is also able to achieve state-of-the-art results but with a much faster inference speed (i.e., it only takes 0.08 second to process a <inline-formula> <tex-math notation="LaTeX">512\times 512 </tex-math></inline-formula> rainy image). Code and pre-trained models are available at <inline-formula> <tex-math notation="LaTeX">\langle </tex-math></inline-formula> https://github.com/RobinCSIRO/AFR-Net <inline-formula> <tex-math notation="LaTeX">\rangle </tex-math></inline-formula>.