CT-SRCNN: Cascade Trained and Trimmed Deep Convolutional Neural Networks for Image Super Resolution

TL;DR

This paper introduces a cascade training and trimming approach for deep CNNs to enhance image super resolution accuracy and efficiency, achieving state-of-the-art results with significantly faster processing times.

Contribution

The paper presents novel cascade training and trimming methodologies to improve CNN accuracy and efficiency for image super resolution.

Findings

Achieves state-of-the-art super resolution accuracy.

Network is over 4 times faster than existing methods.

Maintains discriminative ability despite size reduction.

Abstract

We propose methodologies to train highly accurate and efficient deep convolutional neural networks (CNNs) for image super resolution (SR). A cascade training approach to deep learning is proposed to improve the accuracy of the neural networks while gradually increasing the number of network layers. Next, we explore how to improve the SR efficiency by making the network slimmer. Two methodologies, the one-shot trimming and the cascade trimming, are proposed. With the cascade trimming, the network's size is gradually reduced layer by layer, without significant loss on its discriminative ability. Experiments on benchmark image datasets show that our proposed SR network achieves the state-of-the-art super resolution accuracy, while being more than 4 times faster compared to existing deep super resolution networks.