Single Image Super-Resolution via Cascaded Multi-Scale Cross Network

TL;DR

This paper introduces a cascaded multi-scale cross network for single image super-resolution, enhancing information flow and detail reconstruction through a coarse-to-fine approach with multi-scale modules and residual learning.

Contribution

The paper proposes a novel cascaded multi-scale cross network with residual features learning and cascaded supervision, improving super-resolution performance over existing methods.

Findings

Outperforms state-of-the-art super-resolution methods on benchmark datasets.

Effectively captures multi-scale contextual information for detailed reconstruction.

Demonstrates superior quantitative and qualitative results.

Abstract

The deep convolutional neural networks have achieved significant improvements in accuracy and speed for single image super-resolution. However, as the depth of network grows, the information flow is weakened and the training becomes harder and harder. On the other hand, most of the models adopt a single-stream structure with which integrating complementary contextual information under different receptive fields is difficult. To improve information flow and to capture sufficient knowledge for reconstructing the high-frequency details, we propose a cascaded multi-scale cross network (CMSC) in which a sequence of subnetworks is cascaded to infer high resolution features in a coarse-to-fine manner. In each cascaded subnetwork, we stack multiple multi-scale cross (MSC) modules to fuse complementary multi-scale information in an efficient way as well as to improve information flow across the layers. Meanwhile, by introducing residual-features learning in each stage, the relative information between high-resolution and low-resolution features is fully utilized to further boost reconstruction performance. We train the proposed network with cascaded-supervision and then assemble the intermediate predictions of the cascade to achieve high quality image reconstruction. Extensive quantitative and qualitative evaluations on benchmark datasets illustrate the superiority of our proposed method over state-of-the-art super-resolution methods.