Cross-Modal Stitching is an image stitching framework that integrates spatially aligned inputs from distinct sensing modalities by differentiating through modality-specific feature representations to achieve a unified, coherent composite.