UCAN: Unified Convolutional Attention Network for Expansive Receptive Fields in Lightweight Super-Resolution
This work addresses the need for efficient super-resolution models for deployment on resource-constrained devices, offering an incremental improvement in balancing accuracy and computational cost.
The paper tackles the problem of high computational cost in hybrid CNN-Transformer architectures for image super-resolution by proposing UCAN, a lightweight network that unifies convolution and attention to expand receptive fields efficiently, achieving 31.63 dB PSNR on Manga109 with only 48.4G MACs and outperforming larger models on benchmarks like BSDS100.
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.