论文导读：基于退化引导的空间自适应图像复原

Purohit K, Suin M, Rajagopalan A N, et al. Spatially-adaptive image restoration using distortion-guided networks[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision. 2021: 2309-2319.

实验效果图：

Rain-drop Removal

 Shadow Removal

解决什么问题？

•Prior approaches are typically degradation-specific and employ the same processing across different images anddifferent pixels within.

•

•SPAIR, a network design that harnesses distortion-localization information and dynamically adjusts computation to difficult regions in the image.

 SPAIR结构

(1)A localization network that identifies degraded pixels.

(2) A restoration network that exploits knowledge from the localization network in filter and feature domain to selectively and adaptively restore degraded pixels

 The architecture of SPAIR.

核心模块

1. Spatial Feature Modulator (SFM)

- Efficacy:  SFM is to perform distortion-guided spatial-varying feature normalization.

- Process:  Given the fused features   and the predicted mask   , we calculate the modulated features.

 

 

- Studies(Justin Johnson et al.2016) show the correspondence between

• Feature mean → global semantic information.
• Feature variance → local texture.

- SFM modulates features at M = 1 to match the feature statistics of features at M = 0.

2 Mask-guided Sparse Convolution (SC)

- Efficacy:   Facilitates selective restoration of highly degraded regions, and simplifies the learning process.

- Process:   Given the fused features  and the predicted mask , we calculate the sparse features     .

 Actually, means that->