Image Super-Resolution via Dual-State Recurrent Networks

TL;DR

This paper introduces a dual-state recurrent neural network for image super-resolution that jointly exploits low- and high-resolution signals, improving accuracy and memory efficiency over existing single-state models.

Contribution

The paper proposes the Dual-State Recurrent Network (DSRN), a novel architecture that exchanges information between low- and high-resolution states for enhanced super-resolution performance.

Findings

DSRN outperforms state-of-the-art methods in accuracy.

DSRN uses less memory than comparable models.

Qualitative results show improved image quality.

Abstract

Advances in image super-resolution (SR) have recently benefited significantly from rapid developments in deep neural networks. Inspired by these recent discoveries, we note that many state-of-the-art deep SR architectures can be reformulated as a single-state recurrent neural network (RNN) with finite unfoldings. In this paper, we explore new structures for SR based on this compact RNN view, leading us to a dual-state design, the Dual-State Recurrent Network (DSRN). Compared to its single state counterparts that operate at a fixed spatial resolution, DSRN exploits both low-resolution (LR) and high-resolution (HR) signals jointly. Recurrent signals are exchanged between these states in both directions (both LR to HR and HR to LR) via delayed feedback. Extensive quantitative and qualitative evaluations on benchmark datasets and on a recent challenge demonstrate that the proposed DSRN performs favorably against state-of-the-art algorithms in terms of both memory consumption and predictive accuracy.