Simon-SR: Spatially Adaptive Modulation and Visual Prompt Adaptation for Text-Reinforced Super-Resolution
Simon-SR is a new super-resolution method that uses learnable text prompts to guide image reconstruction from low-resolution inputs. It combines Contrastive Prompt Learning to adapt textual semantics and Prompt-Guided Spatially Adaptive Refinement to fuse visual and text features. This allows the model to create high-quality images by treating text as a flexible, learned variable rather than relying on fixed annotations.