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/
Figures & tables
Figure 1: SemLayer can convert web-sourced icon into designer-friendly, semantically layered representations, enabling intuitive editing (bottom row) that would otherwise be difficult (top row).
Figure 2: Overview of our SemLayer pipeline.
Segmentation Model
Segmentation
Completion
mIoU (%) ↑
PQ (%) ↑
mIoU Refined (%) ↑
PQ Refined (%) ↑
mIoU (%) ↑
CD ↓
gpt-image-1 [ 37 ]
25.4
6.20
57.2
39.3
60.9
71.4
SAM2 [ 42 ]
51.1
26.2
62.2
37.8
69.2
61.7
SAM2 ∗
79.3
59.4
85.3
78.0
80.7
49.1
Ours
84.3
76.1
86.4
78.3
85.2
46.6
Table 1: Comparison of different segmentation models. Segmentation results are reported for both the Original predictions and the Refined predictions, where refinement assigns each unlabeled (black) pixel the color of its nearest labeled pixel.
Completion Model
mIoU (%) ↑
CD (pix) ↓
gpt-image-1
10.7
98.6
MP3D [ 66 ]
70.5
79.4
MP3D-finetuned [ 66 ]
75.3
68.9
Ours
85.2
46.6
Table 2: Comparison of different completion models.
Figure 3: Qualitative comparison of segmentation quality.
Figure 4: Qualitative comparison of completion results.
Figure 5: Progressive outputs from SemLayer. The left four columns show: (1) the input icon, (2) results from our segmentation model, (3) the completed and vectorized output produced by the completion model with layer ordering, and (4) a 2.5D illustration visualizing separated layers. The right two columns present the reference layered SVG and its corresponding reference layer decomposition.
Figure 6: SemLayer recovers a semantic-aware layered representation, enabling flexible part-level understanding, editing and animation that are difficult with the original icons, as shown in the figure.
Figure 7: Illustration of our curve reuse and local surgery pipeline. We (1) detect the contact region between the original contour and the completed contour, (2) cut both curves locally, (3) keep the high-quality original segments (blue), (4) reuse the reliable part of the completion (red), and (5) connect them using newly constructed smooth bridge curves (green).
Online learning platforms are increasingly deploying AI assistants to provide personalized, on-demand tutoring. These can leverage infographics to communicate complex information and improve the learning experience. However, automating effective infographic generation remains a challenging task, requiring a combination of visual richness with precise typography, structured layouts, and continued editability. Existing representation methods satisfy these goals only in part: Scalable Vector Graphics (SVG) representations preserve geometry, hierarchy, and text but provide limited aesthetic flexibility, while raster diffusion models produce richer appearances by flattening content and layout into a single image. Conditioning a diffusion model on a rasterized SVG does not resolve this conflict because the raster preserves approximate placement but discards element ownership, occlusion relationships, and the structure needed to modify and recompose the result. We introduce LayerCanvas, a hybrid approach that retains SVG as the semantic, structural, and typographic program while using diffusion to aestheticize non-text content. Rather than generating one flattened raster, LayerCanvas partitions SVG elements into z-ordered sets of mutually non-overlapping elements and processes them through communicating parallel RGB generation streams. The resulting layers are independently editable and deterministically recomposed, while text is rendered from preserved vector metadata. Our approach combines rich visual synthesis with accurate typography and localized editing. LayerCanvas substantially improves text fidelity over state-of-the-art raster image generators while achieving the highest overall aesthetic quality. We further demonstrate non-destructive content editing, localization, asset reuse, and animation.
Keshav Gupta, Jeremy Owen, Christopher Lloyd +2
University of California, San Diego · LinkedIn Inc.
Recent generative image editing methods adopt layered representations to mitigate the entangled nature of raster images and improve controllability, typically relying on object-based segmentation. However, such strategies may fail to capture the structural and stylized properties of human-created images, such as anime illustrations. To solve this issue, we propose a workflow-aware structured layer decomposition framework tailored to the illustration production of anime artwork. Inspired by the creation pipeline of anime production, our method decomposes the illustration into semantically meaningful production layers, including line art, flat color, shadow, and highlight. To decouple all these layers, we introduce lightweight layer semantic embeddings to provide specific task guidance for each layer. Furthermore, a set of layer-wise losses is incorporated to supervise the training process of individual layers. To overcome the lack of ground-truth layered data, we construct a high-quality illustration dataset that simulated the standard anime production workflow. Experiments demonstrate that the accurate and visually coherent layer decompositions were achieved by using our method. We believe that the resulting layered representation further enables downstream tasks such as recoloring and embedding texture, supporting content creation, and illustration editing. Code is available at: https://github.com/zty0304/Anime-layer-decomposition
Tianyu Zhang, Dongchi Li, Keiichi Sawada +1
Japan Advanced Institute of Science and Technology · Live2D Inc. · Waseda University
We introduce UniWorld-Design, a framework that redefines image generation from flat pixel synthesis to structured visual composition, with semantic RGBA layers as the atomic units of generation, understanding, and editing. Our key insight is that pixels define how an image is rendered, whereas layers define how an image is created, understood, and edited. Just as human designers create and manipulate visual content through layers rather than raw pixels, UniWorld-Design equips multimodal generative models with a layer-native design space. UniWorld-Design comprises two models. The Text-to-RGBA (T2RGBA) model generates standalone RGBA assets directly from text. The Image-to-Layer (I2L) model conditions on a finished image, a global instruction and per-layer prompts, and jointly produces ordered, complete semantic RGBA layers. Its instruction interface supports top-level decomposition, recursive decomposition and targeted extraction, making layering an instruction-addressable operation for agentic editing. Because I2L learns complete semantic objects rather than visible-pixel partitions, its layers stay usable when moved or removed. On the Crello benchmark, I2L reduces per-layer RGB L1 error by 37% and achieves a 34% relative improvement in Alpha Soft IoU over Qwen-Image-Layered. Separately, T2RGBA achieves the highest CLIP Score, outperforming LayerDiffuse and OmniAlpha.