“Two-Stage Sketch-Based Smoke Illustration Generation Using Stream Function” by Chang, Xie, Sato and Xie
Conference:
Type(s):
Title:
- Two-Stage Sketch-Based Smoke Illustration Generation Using Stream Function
Session/Category Title:
- Images, Video & Computer Vision
Presenter(s)/Author(s):
Abstract:
We propose a two-stage sketch-guided smoke illustration generation framework using stream function. The input sketch is converted into the stream function through a latent diffusion model, which subsequently drives the velocity field generation. The velocity field serves as a guidance force to drive the smoke simulation.
References:
[1] Hengyuan Chang, Tianyu Zhang, Syuhei Sato, and Haoran Xie. 2025. DiffSmoke: Two-Stage Sketch-Based Smoke Illustration Design Using Diffusion Models. IEEE Access 13 (2025), 44997–45009.
[2] Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Björn Ommer. 2022. High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 10684–10695.
[3] Syuhei Sato, Yoshinori Dobashi, and Theodore Kim. 2021. Stream-guided smoke simulations. ACM Trans. Graph. 40, 4, Article 161 (July 2021), 7 pages.
[4] Haoran Xie, Keisuke Arihara, Syuhei Sato, and Kazunori Miyata. 2024. Dualsmoke: Sketch-based smoke illustration design with two-stage generative model. Computational Visual Media (2024), 1–15.
