Single Image Super-resolution Based On Self-calibrated Attention Mechanism

Image resolution is a set of performance parameters used to evaluate the richness of detailed information contained in an image,including temporal resolution,spatial resolution and color level resolution,which reflects the ability of the imaging system to reflect the actual detail information of the object.High-resolution images usually contain greater pixel density,richer texture detail,and higher reliability than low-resolution images.Deep Convolutional Neural Networks(DCNNs)have achieved remarkable performance in single image super-resolution(SISR).However,most SR methods restore high resolution(HR)images from single-scale region in the low resolution(LR)input,which limits the ability of method to infer multi-scales of details for high resolution(HR)output.In this paper a novel basic building block called self-Calibrated residual block(SARB)is proposed to solve this problem.SARB consists of carefully designed multi-scale paths,which can capture rich structure information from different scale.In addition,self-Calibrated residual block is introduced to adaptively learn informatively context to make network generate more discriminative representations.These blocks are composed of our self-calibrated attention residual network(SARN)for image super-resolution.Experiments results on five benchmark datasets demonstrate that the proposed SARN achieves comparable results compared with the previous most of the state-of-the-art methodsNext,the super-resolution reconstruction of single image based on residual network of self-correcting attention mechanism is improved in two aspects.The first method is to improve the multi-scale convolution layer into a dense cross-connection layer.The second method is to improve the network structure into a hierarchical residuals attention network.The optimized experimental results are obtained by these two improved methods.