Abstraction in Style DOI: XXXXXXX.XXXXXXX ISBN: 978-1-4503-XXXX-X/2025/XX 430
Min Lu, Yuanfeng He, Anthony Chen, Jianhuang He, Pu Wang, Daniel Cohen-Or, Hui Huang* · 2026
Read Abstraction in Style DOI: XXXXXXX.XXXXXXX ISBN: 978-1-4503-XXXX-X/2025/XX 430 by Min Lu, Yuanfeng He, Anthony Chen, Jianhuang He, Pu Wang, Daniel Cohen-Or, Hui Huang* (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-29924.txt.
Introduction
Lu and co-authors present Abstraction in Style, a method for artistic style transfer. Real styles rework structure, not just texture and color. Older transfer methods keep the input geometry, so they miss that deeper change. Their AiS first builds an abstraction proxy from a target image and style samples. The proxy recasts structure while loosening geometric fidelity. A second stage renders the proxy into the final image. Both stages learn from visual examples alone.
Catalog record
| CREATOR | Min Lu, Yuanfeng He, Anthony Chen, Jianhuang He, Pu Wang, Daniel Cohen-Or, Hui Huang* |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 42009 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.29924
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Abstraction in Style DOI: XXXXXXX.XXXXXXX ISBN: 978-1-4503-XXXX-X/2025/XX 430 Min Lu, Yuanfeng He, Anthony Chen, Jianhuang He, Pu Wang, Daniel Cohen-Or, Hui Huang* Abstract . Artistic styles often embed abstraction beyond surface appearance, involving deliberate reinterpretation of structure rather than mere changes in texture or color. Conventional style transfer methods typically preserve the input geometry and therefore struggle to capture this deeper abstraction behavior, especially for illustrative and non-photorealistic styles. In this work, we introduce Abstraction in Style (AiS) , a generative framework that separates structural abstraction from visual stylization . Given a target image and a small set of style exemplars, AiS first derives an intermediate abstraction proxy that reinterprets the target’s structure in accordance with the abstraction logic exhibited by the style. The proxy captures semantic structure while relaxing geometric fidelity, enabling subsequent stylization to operate on an abstracted representation rather than the original image. In a second stage, the abstraction proxy is rendered to produce the final stylized output, preserving visual coherence with the reference style. Both stages are implemented using a shared image-space analogy, enabling transformations to be learned from visual exemplars without explicit geometric supervision. By decoupling abstraction from appearance and treating abstraction as an explicit, transferable process, AiS supports a wider range of stylistic transformations, improves controllability, and enables more expressive stylization. Figure 1. Abstraction in Style (AiS): Across four challenging illustrative styles with strong structural abstraction, AiS generates coherent results for target images from a small set of exemplars. All reference images in this work sourced from Pinterest © original authors. 1. Introduction The generation of stylized visual content is a longstanding goal in computer graphics ( 10 ) . A substantial body of research is dedicated to style transfer, in which the visual traits of a reference artwork, such as stroke texture, color palette, or ornamentation, are applied to a target image while preserving its general underlying subject. Many works transfer style in a way that adheres to the original spatial structure ( 46 ; 39 ) , even when the target style is abstract, and therefore tend to overlook the abstraction function that may be latent in the reference style. In this paper, we introduce Abstraction in Style (AiS) , a generative framework grounded in the premise that style and abstraction are distinct yet intertwined dimensions of visual expression (see the references in Figure 1 ). While style governs surface appearance, such as stroke quality, texture, or color—abstraction involves a reinterpretation of structure, focusing on re-expressing essential meaning or relationships within forms rather than replicating exact, realistic geometry. Abstraction often alters the geometric form of a subject ( 13 ) : lines may become deliberately irregular ( 41 ) , less significant geometric features can be smoothed away ( 24 ; 11 ) , symmetry may be intentionally broken, and proportions distorted to convey a particular visual character or simplification. These operations reflect a higher-level visual reasoning that conventional style transfer methods, which largely preserve input geometry, are not designed to capture. AiS addresses this limitation by explicitly separating structural abstraction from visual stylization , enabling the model not only to transfer visual appearance but also to capture the abstraction behavior inherent in a reference style, defined by a small set of exemplars. As illustrated in Figure 2 , this explicit treatment of abstraction allows structural reinterpretation beyond rigid geometry preservation, which is essential for many illustrative and non-photorealistic styles. Figure 2. Conventional style transfer keeps the structure of the target image. Introducing an abstraction stage prior to the stylization, allows adopting structural attributes from the reference, leading to a more coherent transformation. Conceptually, AiS formulates stylized image generation as a two-stage process that separates structural abstraction from visual stylization. In the first stage, Structural Abstraction , the structure of the target image is reinterpreted in the spirit of the reference exemplars, producing an intermediate abstraction proxy that reflects the abstraction tendencies in the references while preserving the subject identity of the target. In the second stage, Visual Stylization , this proxy is rendered into a final image that is visually coherent with the reference style. In this paper, we present one concrete realization of this framework. Taking the abstraction proxy as a bridge, both the abstraction and stylization stages are implemented using a shared image-space analogy mechanism, referred to as Visual Analogy Transfer (VAT) . Trained on a small set of style exemplars (averagely 10 per style), VAT learns transformations directly from these images without explicit geometric supervision. While alternative realizations of AiS are possible, our implementation demonstrates that explicitly modeling abstraction alongside appearance yields a more interpretable and controllable stylization process and supports a broader range of structural transformations than conventional style transfer methods. We demonstrate the effectiveness of Abstraction in Style through extensive qualitative results and evaluations. As shown in Figure 1 , AiS supports stylized image generation across diverse illustrative styles, producing outputs that reflect both the visual appearance and the underlying abstraction logic of the reference. By explicitly modeling abstraction as a separate stage, AiS yields more coherent structural reinterpretations than conventional style transfer methods, particularly for non-photorealistic and highly abstract styles. Comparative experiments, ablations, and user studies further validate that disentangling abstraction from appearance improves stylistic fidelity, structural expressiveness, and controllability. 2. Related Work Prior work related to Abstraction in Style spans two largely independent directions: visual abstraction, which reinterprets structure, and style transfer, which modifies appearance while typically preserving geometry. While both have been extensively studied, they have largely evolved separately, without explicitly modeling abstraction as a transferable process conditioned on style exemplars. Visual abstraction has long been a core concept in art and design ( 13 ) . Early computational approaches explored geometry-driven abstractions through skeletal representations, part-based models, planar slicing, and shape proxies ( 6 ; 8 ; 23 ; 25 ; 30 ; 24 ; 27 ; 42 ) . In imagery, abstraction is closely tied to non-photorealistic rendering ( 7 ) , with seminal work on painterly and brush-based representations ( 12 ; 16 ; 21 ; 32 ) . Sketch-based abstraction has emerged as a particularly expressive domain, including data-driven stroke reassembly and learning-based stroke pruning approaches ( 2 ; 26 ; 22 ; 38 ; 37 ) . More recent optimization-based methods leverage pretrained vision–language models to achieve controllable semantic abstraction without supervised sketch data ( 38 ; 37 ) . While effective at producing abstract representations, these approaches typically operate in a fixed or implicit style and do not condition abstraction behavior on exemplars or support stylization as a separate controllable dimension. Style transfer aims to modify visual appearance while preserving underlying structure, most notably introduced by Gatys et al. ( 10 ) . Subsequent work improved efficiency and flexibility through feed-forward networks, feature normalization, and universal style representations ( 19 ; 35 ; 36 ; 18 ; 3 ; 20 ; 45 ) . Recent diffusion-based methods enable exemplar-based stylization by modulating attention or adapting pretrained models with lightweight modules, allowing consistent appearance transfer across generations ( 1 ; 14 ; 5 ; 9 ; 31 ; 28 ; 29 ) . Despite their effectiveness, these methods largely preserve the input geometry and do not explicitly model