arXiv subjects

Hengyuan Chang

Publications and source records attributed to Hengyuan Chang.

3 recordsLinked to original sources

arXiv2026

Sketch-Guided Stylized Landscape Cinemagraph Synthesis

Designing stylized cinemagraphs is challenging due to the difficulty in customizing complex and expressive flow elements. To achieve intuitive and detailed control of the generated cinemagraphs, sketches provide a feasible solution to convey personalized design requirements beyond text inputs. In this paper, we propose Sketch2Cinemagraph, a sketch-guided framework that enables the conditional generation of stylized cinemagraphs from freehand sketches. Sketch2Cinemagraph adopts text prompts for initial landscape generation and provides sketch controls for both spatial and motion cues. The latent diffusion model first generates target stylized landscape images along with realistic versions. Then, a pre-trained object detection model obtains masks for the flow regions. We propose a latent motion diffusion model to estimate motion field in fluid regions of the generated landscape images. The input motion sketches serve as the conditions to control the generated motion fields in the masked fluid regions with the prompt. To synthesize cinemagraph frames, the pixels within fluid regions are warped to target locations at each timestep using a U-Net based frame generator. The results verified that Sketch2Cinemagraph can generate aesthetically appealing stylized cinemagraphs with continuous temporal flow from sketch inputs. We showcase the advantages of Sketch2Cinemagraph through qualitative and quantitative comparisons against the state-of-the-art approaches.

cs.CV↗

arXiv2025

Two-Stage Sketch-Based Smoke Illustration Generation using Stream Function

In this paper, we propose a two-stage sketch-based smoke illustration generation framework using stream function and latent diffusion models (LDM). The user sketch is used to guide the generation of the stream function, which serves as the control condition for the velocity field generator. The generated velocity field can be used to guide the smoke simulation to align with the intended flow. We adopt streamlines to encode global flow dynamics as sketch guidance during training. The stream function constitutes the intermediate representation that captures continuous variation and rotational flow details absent from sketches.

cs.GR↗

arXiv2025

Physics-Aware Fluid Field Generation from User Sketches Using Helmholtz-Hodge Decomposition

Fluid simulation techniques are widely used in various fields such as film production, but controlling complex fluid behaviors remains challenging. While recent generative models enable intuitive generation of vector fields from user sketches, they struggle to maintain physical properties such as incompressibility. To address these issues, this paper proposes a method for interactively designing 2D vector fields. Conventional generative models can intuitively generate vector fields from user sketches, but remain difficult to consider physical properties. Therefore, we add a simple editing process after generating the vector field. In the first stage, we use a latent diffusion model~(LDM) to automatically generate initial 2D vector fields from user sketches. In the second stage, we apply the Helmholtz-Hodge decomposition to locally extract physical properties such as incompressibility from the results generated by LDM and recompose them according to user intentions. Through multiple experiments, we demonstrate the effectiveness of our proposed method.

cs.GR↗