SemLayer: Semantic-aware Generative Segmentation and Layer Construction for Abstract Icons
This addresses a domain-specific problem for graphic designers and animators by enabling semantic decomposition of flattened vector art, though it is incremental as it builds on existing visual generation techniques.
The paper tackles the problem of reconstructing semantic layers from flattened vector icons, which hinders editing and animation, by introducing SemLayer, a pipeline that generates visually separable components and recovers complete geometries, enabling previously inapplicable editing workflows.
Graphic icons are a cornerstone of modern design workflows, yet they are often distributed as flattened single-path or compound-path graphics, where the original semantic layering is lost. This absence of semantic decomposition hinders downstream tasks such as editing, restyling, and animation. We formalize this problem as semantic layer construction for flattened vector art and introduce SemLayer, a visual generation empowered pipeline that restores editable layered structures. Given an abstract icon, SemLayer first generates a chromatically differentiated representation in which distinct semantic components become visually separable. To recover the complete geometry of each part, including occluded regions, we then perform a semantic completion step that reconstructs coherent object-level shapes. Finally, the recovered parts are assembled into a layered vector representation with inferred occlusion relationships. Extensive qualitative comparisons and quantitative evaluations demonstrate the effectiveness of SemLayer, enabling editing workflows previously inapplicable to flattened vector graphics and establishing semantic layer reconstruction as a practical and valuable task. Project page: https://xxuhaiyang.github.io/SemLayer/