Sebica: Lightweight Spatial and Efficient Bidirectional Channel Attention Super Resolution Network
This work addresses the problem of deploying high-quality SISR models efficiently for applications like traffic video object detection, though it is incremental as it builds on existing lightweight attention mechanisms.
The paper tackles the computational challenges of single image super-resolution (SISR) in resource-limited environments by introducing Sebica, a lightweight network that reduces parameters and GFLOPs by up to 97% while achieving PSNR/SSIM scores of up to 30.18/0.8330 on datasets like Div2K and Flickr2K.
Single Image Super-Resolution (SISR) is a vital technique for improving the visual quality of low-resolution images. While recent deep learning models have made significant advancements in SISR, they often encounter computational challenges that hinder their deployment in resource-limited or time-sensitive environments. To overcome these issues, we present Sebica, a lightweight network that incorporates spatial and efficient bidirectional channel attention mechanisms. Sebica significantly reduces computational costs while maintaining high reconstruction quality, achieving PSNR/SSIM scores of 28.29/0.7976 and 30.18/0.8330 on the Div2K and Flickr2K datasets, respectively. These results surpass most baseline lightweight models and are comparable to the highest-performing model, but with only 17% and 15% of the parameters and GFLOPs. Additionally, our small version of Sebica has only 7.9K parameters and 0.41 GFLOPS, representing just 3% of the parameters and GFLOPs of the highest-performing model, while still achieving PSNR and SSIM metrics of 28.12/0.7931 and 0.3009/0.8317, on the Flickr2K dataset respectively. In addition, Sebica demonstrates significant improvements in real-world applications, specifically in object detection tasks, where it enhances detection accuracy in traffic video scenarios.