Sebica: Lightweight Spatial and Efficient Bidirectional Channel Attention Super Resolution Network

TL;DR

Sebica is a lightweight super-resolution network that efficiently balances high image reconstruction quality with low computational costs, making it suitable for resource-constrained environments and improving real-world object detection.

Contribution

We introduce Sebica, a novel lightweight SISR model with spatial and bidirectional channel attention, achieving high performance with significantly fewer parameters and GFLOPs.

Findings

Achieves PSNR/SSIM of 28.29/0.7976 on Div2K dataset.

Small version with 7.9K parameters still performs well.

Enhances object detection accuracy in traffic videos.

Abstract

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.