Cora: Correspondence-aware image editing using few step diffusion
This addresses image editing challenges in computer graphics, vision, and VFX for users needing precise structural modifications, representing a novel method for a known bottleneck rather than incremental.
The paper tackles the problem of image editing requiring structural changes like non-rigid deformations, where existing few-step diffusion methods produce artifacts or fail to preserve key attributes. The result is Cora, a framework that uses correspondence-aware noise correction and interpolated attention maps to maintain structure, textures, and identity across diverse edits, with user studies confirming superior performance.
Image editing is an important task in computer graphics, vision, and VFX, with recent diffusion-based methods achieving fast and high-quality results. However, edits requiring significant structural changes, such as non-rigid deformations, object modifications, or content generation, remain challenging. Existing few step editing approaches produce artifacts such as irrelevant texture or struggle to preserve key attributes of the source image (e.g., pose). We introduce Cora, a novel editing framework that addresses these limitations by introducing correspondence-aware noise correction and interpolated attention maps. Our method aligns textures and structures between the source and target images through semantic correspondence, enabling accurate texture transfer while generating new content when necessary. Cora offers control over the balance between content generation and preservation. Extensive experiments demonstrate that, quantitatively and qualitatively, Cora excels in maintaining structure, textures, and identity across diverse edits, including pose changes, object addition, and texture refinements. User studies confirm that Cora delivers superior results, outperforming alternatives.