 |
SSILK: Self-Supervised Integration of Latent Kinematics for Joint-Driven Neural GarmentsMaksym Perepichka, Arnaud Schoentgen, Eric Paquette, and Tiberiu Popa, Computer Graphics Forum, Vol. 45, Issue 8, 2026. Presented at the Symposium on Computer Animation (SCA), Barcelola, Spain, July 8-10, 2026 |
Garment animation systems are essential for high-fidelity digital characters in various applications such as films and video games. We propose SSILK, a neural network-based method that is explicitly designed to handle loose garments such as capes and robes, garment types that are notoriously hard to simulate via pose-based garment neural simulators. We achieve this through three contributions. First, we formulate our method as a latent-space time integration scheme inspired by Verlet integration, resulting in more temporally stable garment dynamics compared to prior methods. Second, we employ a hybrid proxy-joint and blendshape decoder to represent full SO(3) transformations that can be too challenging to capture using blendshapes alone. Finally, we propose a data-augmentation strategy to improve performance on out-of-distribution scenarios. We demonstrate the effectiveness of our method on various garment and character animations and compare it to the state of the art.
cloth simulation, neural networks, physics-based learning, character animation, deep learning
@article{Perepichka2026,
journal = {Computer Graphics Forum},
title = {{SSILK : Self-Supervised Integration of Latent Kinematics for Joint-Driven Neural Garments}},
author = {Perepichka, Maksym and Schoentgen, Arnaud and Paquette, Eric and Popa, Tiberiu},
year = {2026},
publisher = {The Eurographics Association},
ISSN = {1467-8659},
DOI = {10.1111/cgf.70579}
}
Official published paper: https://doi.org/10.1111/cgf.70579.
Preliminary version of the paper.
Slides of the presentation.
Pre-print version of the video:
