Channel-Partitioned Windowed Attention And Frequency Learning for Single Image Super-Resolution
This addresses the problem of enhancing image quality in SISR for applications like photography or medical imaging, though it appears incremental as it builds on existing transformer-based methods.
The paper tackled the limitations of window-based attention in capturing long-range dependencies and frequency content in single image super-resolution (SISR), proposing a Channel-Partitioned Attention Transformer and Spatial-Frequency Interaction Module that achieved up to 0.31dB improvement over state-of-the-art methods on Urban100 at x2 SR.
Recently, window-based attention methods have shown great potential for computer vision tasks, particularly in Single Image Super-Resolution (SISR). However, it may fall short in capturing long-range dependencies and relationships between distant tokens. Additionally, we find that learning on spatial domain does not convey the frequency content of the image, which is a crucial aspect in SISR. To tackle these issues, we propose a new Channel-Partitioned Attention Transformer (CPAT) to better capture long-range dependencies by sequentially expanding windows along the height and width of feature maps. In addition, we propose a novel Spatial-Frequency Interaction Module (SFIM), which incorporates information from spatial and frequency domains to provide a more comprehensive information from feature maps. This includes information about the frequency content and enhances the receptive field across the entire image. Experimental findings show the effectiveness of our proposed modules and architecture. In particular, CPAT surpasses current state-of-the-art methods by up to 0.31dB at x2 SR on Urban100.