Multi-task learning for image restoration
Tetiana Martyniuk · 2019
We present an efficient end-to-end pipeline for general image restoration. The setting has a generic encoder and separate decoders so that our model can benefit from the shared low-level feature representations between the tasks. We also introduce the new architecture for the generator inspired by the feature pyramid networks for dealing with multi-scale degradations. We train the models for solving three particular image restoration problems: deblurring, dehazing, and raindrop removal.