DARN: Distance Attention Residual Network for Lightweight Remote-Sensing Image Superresolution

Abstract

The application of single-image superresolution (SISR) in remote sensing is of great significance. Although the state-of-the-art convolution neural network (CNN)-based SISR methods have achieved excellent results, the large model and slow speed make it difficult to deploy in real remote sensing tasks. In this article, we propose a compact and efficient distance attention residual network (DARN) to achieve a better compromise between model accuracy and complexity. The distance attention residual connection block (DARCB), the core component of the DARN, uses multistage feature aggregation to learn more accurate feature representations. The main branch of the DARCB adopts a shallow residual block (SRB) to flexibly learn residual information to ensure the robustness of the model. We also propose a distance attention block (DAB) as a bridge between the main branch and the branch of the DARCB; the DAB can effectively alleviate the loss of detail features in the deep CNN extraction process. Experimental results on two remote sensing and five super-resolution benchmark datasets demonstrate that the DARN achieves a better compromise than existing methods in terms of performance and model complexity. In addition, the DARN achieves the optimal solution compared with the state-of-the-art lightweight remote sensing SISR method in terms of parameter amount, computation amount, and inference speed. Our code will be available at https://github.com/candygogogogo/DARN.

Lightweight interactive feature inference network for single-image super-resolution

“…Both RS-1 and RS-2 contain 120 images and cover diverse scenes with complicated image patterns. We exploit existing remote-sensing SR methods for comparison, including SRCNN 1 , VDSR 15 , LGCNet 65 , LapSRN 39 , IDN 19 , LESRCNN 44 , CARN-M 6 , FENet 63 , FDENet 66 , and DRAN 67 . All the aforementioned methods are directly evaluated on remote sensing data utilizing pre-trained models provided by relevant workers.…”

Decoupled Knowledge Distillation via Spatial Feature Blurring for Hyperspectral Image Classification

“…All along, how to make models lightweight has always been an important research area [60], [61]. We will continue to explore further optimizations of the knowledge distillation strategy to achieve more efficient knowledge transfer, thereby further enhancing the performance of the student network.…”

Dual-Attention-Guided Multiscale Feature Aggregation Network for Remote Sensing Image Change Detection

“…Therefore, in recent years, deep learning methods have gradually become a research hotspot in remote sensing image change detection [19]. The deep learning method can automatically learn the features in the image without manual feature extraction [20][21][22], and has strong nonlinear modeling ability, which can better adapt to the change detection task. From the early proposed fully convolutional neural network (FCNS) [23][24][25], to the typical convolutional neural network (CNNS) [26][27][28], and then to the recently emerging Transformer network [29,30].…”