Deep Inception-Residual Laplacian Pyramid Networks for Accurate Single Image Super-Resolution

TL;DR

This paper introduces a deep inception-residual Laplacian pyramid network for single image super-resolution, employing a novel two-stage training strategy and a combined loss function to enhance high-frequency detail restoration and outperform existing methods.

Contribution

It proposes a new deep network architecture with inception-residual blocks within a Laplacian pyramid framework, along with a two-stage training method and a dual-space loss function for improved super-resolution accuracy.

Findings

Outperforms state-of-the-art SR methods in objective metrics.

Achieves superior visual quality in reconstructed high-resolution images.

Demonstrates effective training of very deep networks with gradient clipping.

Abstract

With exploiting contextual information over large image regions in an efficient way, the deep convolutional neural network has shown an impressive performance for single image super-resolution (SR). In this paper, we propose a deep convolutional network by cascading the well-designed inception-residual blocks within the deep Laplacian pyramid framework to progressively restore the missing high-frequency details of high-resolution (HR) images. By optimizing our network structure, the trainable depth of the proposed network gains a significant improvement, which in turn improves super-resolving accuracy. With our network depth increasing, however, the saturation and degradation of training accuracy continues to be a critical problem. As regard to this, we propose an effective two-stage training strategy, in which we firstly use images downsampled from the ground-truth HR images as the optimal objective to train the inception-residual blocks in each pyramid level with an extremely high learning rate enabled by gradient clipping, and then the ground-truth HR images are used to fine-tune all the pre-trained inception-residual blocks for obtaining the final SR model. Furthermore, we present a new loss function operating in both image space and local rank space to optimize our network for exploiting the contextual information among different output components. Extensive experiments on benchmark datasets validate that the proposed method outperforms existing state-of-the-art SR methods in terms of the objective evaluation as well as the visual quality.