Back-Projection Pipeline

Abstract

We propose a simple extension of residual networks that works simultaneously in multiple resolutions. Our network design is inspired by the iterative back-projection algorithm but seeks the more difficult task of learning how to enhance images. Compared to similar approaches, we propose a novel solution to make back-projections run in multiple resolutions by using a data pipeline workflow. Features are updated at multiple scales in each layer of the network. The update dynamic through these layers includes interactions between different resolutions in a way that is causal in scale, and it is represented by a system of ODEs, as opposed to a single ODE in the case of ResNets. The system can be used as a generic multi-resolution approach to enhance images. We test it on several challenging tasks with special focus on super-resolution and raindrop removal. Our results are competitive with state-of-the-arts and show a strong ability of our system to learn both global and local image features.

References in corpus (8)

• Sketch-based Manga Retrieval using Manga109 Dataset
• Training Deep Nets with Sublinear Memory Cost
• The Renormalization Group and Singular Perturbations: Multiple-Scales, Boundary Layers and Reductive Perturbation Theory
• Wide Activation for Efficient and Accurate Image Super-Resolution
• Multi-Scale Dense Networks for Resource Efficient Image Classification
• Residual Non-local Attention Networks for Image Restoration
• Bridging the Gaps Between Residual Learning, Recurrent Neural Networks and Visual Cortex
• Residual Dense Network for Image Restoration