Partial Large Kernel CNNs for Efficient Super-Resolution
This work addresses efficiency challenges in super-resolution for image processing applications, offering a novel CNN-based approach that outperforms transformers with significant computational savings.
The paper tackles the problem of computational inefficiency in super-resolution by proposing Partial Large Kernel CNNs (PLKSR), which reduce latency by 86% compared to naive large kernels and achieve state-of-the-art performance on four datasets at scale ×4, with 68.1% lower latency and 80.2% lower GPU memory usage than SRFormer-light.
Recently, in the super-resolution (SR) domain, transformers have outperformed CNNs with fewer FLOPs and fewer parameters since they can deal with long-range dependency and adaptively adjust weights based on instance. In this paper, we demonstrate that CNNs, although less focused on in the current SR domain, surpass Transformers in direct efficiency measures. By incorporating the advantages of Transformers into CNNs, we aim to achieve both computational efficiency and enhanced performance. However, using a large kernel in the SR domain, which mainly processes large images, incurs a large computational overhead. To overcome this, we propose novel approaches to employing the large kernel, which can reduce latency by 86\% compared to the naive large kernel, and leverage an Element-wise Attention module to imitate instance-dependent weights. As a result, we introduce Partial Large Kernel CNNs for Efficient Super-Resolution (PLKSR), which achieves state-of-the-art performance on four datasets at a scale of $\times$4, with reductions of 68.1\% in latency and 80.2\% in maximum GPU memory occupancy compared to SRFormer-light.