UCAN: Unified Convolutional Attention Network for Expansive Receptive Fields in Lightweight Super-Resolution

TL;DR

UCAN is a lightweight convolutional attention network that efficiently expands receptive fields for high-quality image super-resolution, balancing accuracy and computational cost.

Contribution

It introduces a unified architecture combining window-based attention and Hedgehog Attention, with cross-layer sharing and distillation modules for efficient high-resolution image restoration.

Findings

UCAN-L achieves 31.63 dB PSNR on Manga109 with 48.4G MACs.

UCAN outperforms recent lightweight models on Manga109 and BSDS100 datasets.

Extensive experiments demonstrate UCAN's superior efficiency and scalability.

Abstract

Hybrid CNN-Transformer architectures achieve strong results in image super-resolution, but scaling attention windows or convolution kernels significantly increases computational cost, limiting deployment on resource-constrained devices. We present UCAN, a lightweight network that unifies convolution and attention to expand the effective receptive field efficiently. UCAN combines window-based spatial attention with a Hedgehog Attention mechanism to model both local texture and long-range dependencies, and introduces a distillation-based large-kernel module to preserve high-frequency structure without heavy computation. In addition, we employ cross-layer parameter sharing to further reduce complexity. On Manga109 ($4\times$), UCAN-L achieves 31.63 dB PSNR with only 48.4G MACs, surpassing recent lightweight models. On BSDS100, UCAN attains 27.79 dB, outperforming methods with significantly larger models. Extensive experiments show that UCAN achieves a superior trade-off between accuracy, efficiency, and scalability, making it well-suited for practical high-resolution image restoration.