Real-time Garment Synthesis for Game Deployment

Cloth animation is essential in video games, but creating realistic cloth movements in real-time is challenging due to computing limitations. Traditional methods use simplified simulations combined with keyframing to manage performance and control, but these can be unstable and offer limited animator flexibility.

Our project proposes a solution using neural networks to animate garments. By leveraging the latest advancements in artificial intelligence, we model cloth dynamics as a complex spatiotemporal behaviour which cannot be modelled by traditional approaches. This involves considering the body’s shape and movement along with the garment’s properties to predict how the cloth should move.

By integrating AI-driven models, our approach promises a significant leap forward in cloth animation technology, offering a more realistic and responsive experience in video games. This innovation could transform the way cloth is animated, setting new standards in the gaming industry

Faculty Supervisor:

Tiberiu Popa

Student:

Partner:

Ubisoft Divertissement

Discipline:

Computer science

Sector:

Information and cultural industries; Manufacturing

University:

Concordia University

Program: