RDRN: Recursively Defined Residual Network for Image Super-Resolution
This work addresses performance bottlenecks in deep learning for image super-resolution, which is important for applications like medical imaging or photography, but it is incremental as it builds on existing attention and residual network methods.
The authors tackled the problem of training difficulty and limited performance gains in very deep convolutional neural networks for single image super-resolution by proposing a recursively defined residual network (RDRN) that efficiently utilizes attention blocks, achieving state-of-the-art results with improvements of up to 0.43 dB on benchmarks.
Deep convolutional neural networks (CNNs) have obtained remarkable performance in single image super-resolution (SISR). However, very deep networks can suffer from training difficulty and hardly achieve further performance gain. There are two main trends to solve that problem: improving the network architecture for better propagation of features through large number of layers and designing an attention mechanism for selecting most informative features. Recent SISR solutions propose advanced attention and self-attention mechanisms. However, constructing a network to use an attention block in the most efficient way is a challenging problem. To address this issue, we propose a general recursively defined residual block (RDRB) for better feature extraction and propagation through network layers. Based on RDRB we designed recursively defined residual network (RDRN), a novel network architecture which utilizes attention blocks efficiently. Extensive experiments show that the proposed model achieves state-of-the-art results on several popular super-resolution benchmarks and outperforms previous methods by up to 0.43 dB.