Training Patch Analysis and Mining Skills for Image Restoration Deep Neural Networks

TL;DR

This paper analyzes the impact of training patch selection on image restoration neural networks and proposes guidelines to improve training data preparation for better reproducibility and performance.

Contribution

It introduces a detailed analysis of training patch extraction methods and offers practical guidelines for preparing training data in image restoration tasks.

Findings

Different patch extraction methods significantly affect network performance

Proposed guidelines improve training data quality and reproducibility

Analysis aids in reducing dataset collection costs

Abstract

There have been numerous image restoration methods based on deep convolutional neural networks (CNNs). However, most of the literature on this topic focused on the network architecture and loss functions, while less detailed on the training methods. Hence, some of the works are not easily reproducible because it is required to know the hidden training skills to obtain the same results. To be specific with the training dataset, few works discussed how to prepare and order the training image patches. Moreover, it requires a high cost to capture new datasets to train a restoration network for the real-world scene. Hence, we believe it is necessary to study the preparation and selection of training data. In this regard, we present an analysis of the training patches and explore the consequences of different patch extraction methods. Eventually, we propose a guideline for the patch extraction from given training images.