Institution profile

Industrial Light & Magic

Industry researchnorthamerica · us
Official website
Research library2linked papers
Opportunities0open roles
Selected work

Representative Papers

Over++: Generative Video Compositing for Layer Interaction Effects

Dec 22, 2025

In professional video compositing, environment interactions—such as shadows, reflections, dust, and splashes—between foreground and background have traditionally relied on labor-intensive manual creation. Existing video generation models struggle to inject photorealistic interactions while preserving the input video; conversely, video inpainting methods suffer from either requiring frame-wise manual masks or producing geometrically distorted outputs. To address this, we introduce “enhanced compositing” as a novel task, construct the first paired dataset of videos with environment interaction effects, and propose a self-supervised, unpaired training strategy jointly guided by text prompts, segmentation masks, and keyframes. Our method integrates video diffusion models, unsupervised spatiotemporal consistency modeling, lightweight mask fusion, and keyframe-guided distillation. Experiments demonstrate that our approach generates diverse, high-fidelity semi-transparent interactions under data constraints, achieving state-of-the-art performance in both interaction realism and source scene fidelity.

0 citationsRead paper

Escher Tile Deformation via Closed-Form Solution

Jun 29, 2025

This work addresses the challenge of achieving gap-free, overlap-free real-time deformation of Escher tiles—seamless organic tiling patterns conforming to all 17 wallpaper groups. We propose a closed-form analytical modeling framework based on periodic displacement fields, enabling unified handling of both boundary and interior deformations. To support semantics-aware local artistic editing, we introduce a tunable adaptive attenuation mechanism. Our method is representation-agnostic, operating seamlessly on both image and mesh inputs, and provides joint global–local interactive control. Extensive experiments demonstrate its effectiveness across image editing, shape carving, digital fabrication, and animation generation. It achieves millisecond-level response times, strictly preserves topology throughout deformation, and—crucially—enables, for the first time, real-time, seamless, and controllable dynamic deformation of Escher tiles under arbitrary wallpaper group symmetries.

0 citationsRead paper
Recent publications

Latest Papers

Over++: Generative Video Compositing for Layer Interaction Effects

Dec 22, 2025

In professional video compositing, environment interactions—such as shadows, reflections, dust, and splashes—between foreground and background have traditionally relied on labor-intensive manual creation. Existing video generation models struggle to inject photorealistic interactions while preserving the input video; conversely, video inpainting methods suffer from either requiring frame-wise manual masks or producing geometrically distorted outputs. To address this, we introduce “enhanced compositing” as a novel task, construct the first paired dataset of videos with environment interaction effects, and propose a self-supervised, unpaired training strategy jointly guided by text prompts, segmentation masks, and keyframes. Our method integrates video diffusion models, unsupervised spatiotemporal consistency modeling, lightweight mask fusion, and keyframe-guided distillation. Experiments demonstrate that our approach generates diverse, high-fidelity semi-transparent interactions under data constraints, achieving state-of-the-art performance in both interaction realism and source scene fidelity.

0 citationsRead paper

Escher Tile Deformation via Closed-Form Solution

Jun 29, 2025

This work addresses the challenge of achieving gap-free, overlap-free real-time deformation of Escher tiles—seamless organic tiling patterns conforming to all 17 wallpaper groups. We propose a closed-form analytical modeling framework based on periodic displacement fields, enabling unified handling of both boundary and interior deformations. To support semantics-aware local artistic editing, we introduce a tunable adaptive attenuation mechanism. Our method is representation-agnostic, operating seamlessly on both image and mesh inputs, and provides joint global–local interactive control. Extensive experiments demonstrate its effectiveness across image editing, shape carving, digital fabrication, and animation generation. It achieves millisecond-level response times, strictly preserves topology throughout deformation, and—crucially—enables, for the first time, real-time, seamless, and controllable dynamic deformation of Escher tiles under arbitrary wallpaper group symmetries.

0 citationsRead paper