Learning Dual Convolutional Neural Networks for Low-Level Vision

TL;DR

This paper introduces a flexible dual CNN framework that simultaneously recovers structures and details for various low-level vision tasks, demonstrating superior performance over existing methods.

Contribution

The paper presents a novel dual CNN architecture that effectively separates and recovers structures and details in low-level vision problems, adaptable to multiple tasks.

Findings

Effective in super-resolution, deraining, dehazing, and edge-preserving filtering

Outperforms state-of-the-art methods in various low-level vision tasks

Flexible framework easily integrated into existing CNN architectures

Abstract

In this paper, we propose a general dual convolutional neural network (DualCNN) for low-level vision problems, e.g., super-resolution, edge-preserving filtering, deraining and dehazing. These problems usually involve the estimation of two components of the target signals: structures and details. Motivated by this, our proposed DualCNN consists of two parallel branches, which respectively recovers the structures and details in an end-to-end manner. The recovered structures and details can generate the target signals according to the formation model for each particular application. The DualCNN is a flexible framework for low-level vision tasks and can be easily incorporated into existing CNNs. Experimental results show that the DualCNN can be effectively applied to numerous low-level vision tasks with favorable performance against the state-of-the-art methods.