Abstract: This study proposes a novel general image fusion framework based on cross-domain long-range learning and Swin Transformer, termed as SwinFusion. On the one hand, an attention-guided ...