← ArchivePaper2016
CAMA: Contact-Aware Matrix Assembly with Unified Collision Handling for GPU-based Cloth Simulation
Min Tang, Huamin Wang, Le Tang, Ruofeng Tong, Dinesh Manocha
GPU cloth pipeline that integrates contact forces into implicit time integration via parallelized sparse matrix assembly for 100-300K triangle garments.
Abstract
We present a novel GPU‐based approach to robustly and efficiently simulate high‐resolution and complexly layered cloth. The key component of our formulation is a parallelized matrix assembly algorithm that can quickly build a large and sparse matrix in a compressed format and accurately solve linear systems on GPUs. We also present a fast and integrated solution for parallel collision handling, including collision detection and response computations, which utilizes spatio‐temporal coherence. We combine these algorithms as part of a new cloth simulation pipeline that incorporates contact forces into implicit time integration for collision avoidance. The entire pipeline is implemented on GPUs, and we evaluate its performance on complex benchmarks consisting of 100, 300K triangles. In practice, our system takes a few seconds to simulate one frame of a complex cloth scene, which represents significant speedups over prior CPU and GPU‐based cloth simulation systems.
How to read this
- Category
- Method: a GPU cloth-simulation pipeline with unified collision handling
- Contributions
- Introduces a parallelized contact-aware matrix assembly that builds a large sparse matrix in compressed format and solves the linear system on the GPU
- Presents an integrated parallel collision-handling solution (detection plus response) exploiting spatio-temporal coherence
- Combines these into a pipeline that folds contact forces into implicit time integration for collision avoidance, evaluated on 100-300K-triangle layered cloth
- Context
- Advances robust simultaneous-collision treatment in cloth (e.g. Harmon et al., harmon-simultaneous-collisions-2008) by mapping contact-aware implicit integration onto the GPU.Builds on: Robust Treatment of Simultaneous Collisions
- Correctness
- The contributions target throughput on high-resolution layered garments and the paper reports per-frame times of a few seconds with speedups over prior CPU/GPU systems; robustness claims rest on the unified collision handling, so results should be read as performance-and-robustness on the stated benchmark scale rather than guaranteed for arbitrary scenes.
- Clarity
- Systems-and-numerics heavy; a first pass conveys the pipeline and where the speedups come from, a second pass is required for the matrix-assembly and collision details.
- How to read it
- Read first for the pipeline structure and the contact-aware-integration idea; a second and third pass pay off for the GPU sparse-assembly and parallel collision algorithms if you build cloth solvers.
Builds on
Built upon by
Related work
- Robust Treatment of Collisions, Contact and Friction for a Skinned Cloth Simulation 2002 / SIGGRAPH
- A Safe and Fast Repulsion Method for GPU-based Cloth Self Collisions 2020 / TOG
- Better Collisions and Faster Cloth for Pixar's Coco 2018 / SIGGRAPH
- Smoothed Aggregation Multigrid for Cloth Simulation 2015 / SIGGRAPH Asia
Keywords
This page summarises the entry and links to its original source. The archive never hosts or redistributes the publication itself.Show it in the full archive list →