# Awesome Deformable Object Simulation for Embodied AI 🧸 [![Awesome](https://awesome.re/badge.svg)](https://awesome.re) > A curated list of resources on simulating deformable objects (cloth, rope, soft bodies) for embodied intelligence β€” simulation environments, manipulation tasks, sim-to-real transfer, and asset generation. [Project website](https://codeofwhite.github.io/awesome-deformable-manipulation/) | [English](README.md) | [δΈ­ζ–‡](README.zh-CN.md) --- ## Contents - [Simulation Environments & Platforms](#simulation-environments--platforms) - [Deformable Object Manipulation Tasks](#deformable-object-manipulation-tasks) - [Sim-to-Real Transfer](#sim-to-real-transfer) - [Benchmarks & Evaluation](#benchmarks--evaluation) - [3D Asset Generation for Simulation](#3d-asset-generation-for-simulation) - [Foundational Simulation Techniques](#foundational-simulation-techniques) - [Legend](#legend) --- ## Simulation Environments & Platforms - [SAPIEN: A SimulAted Part-based Interactive ENvironment](https://arxiv.org/abs/2003.08515) - Interactive simulation environment based on PhysX, excelling at articulated and deformable object simulation. Widely used in Embodied AI research. πŸ“„ πŸ”§ ⭐ - [Genesis World](https://genesis-world.readthedocs.io/) - Multi-physics simulation platform covering rigid, deformable, and fluid bodies. 29k+ GitHub stars. πŸ”§ ⭐ - [SAPIEN ManiSkill](https://maniskill2.github.io/) - Unified framework for manipulation skill learning built on SAPIEN, with rich deformable object tasks. πŸ”§ πŸ“Š ⭐ - [Isaac Sim / Isaac Lab](https://developer.nvidia.com/isaac-sim) - NVIDIA's robotics simulation platform with GPU-accelerated deformable body support for large-scale Embodied AI training. πŸ”§ ⭐ - [MuJoCo](https://mujoco.org/) - Advanced physics engine with deformable body simulation support (since v3.x), widely adopted in robotics research. πŸ”§ ⭐ - [Taichi](https://github.com/taichi-dev/taichi) - High-performance parallel computing language, popular for differentiable soft body simulation in Embodied AI research. πŸ”§ - [Real2Render2Real: Scaling Robot Data Without Dynamics Simulation or Robot Hardware](https://arxiv.org/abs/2505.09601) - Scales robot training data via a Realβ†’Renderβ†’Real pipeline without dynamics simulation or physical robot hardware. πŸ“„ ⭐ ## Deformable Object Manipulation Tasks - [GarmentLab: A Unified Simulation and Benchmark for Garment Manipulation](https://arxiv.org/abs/2411.01200) - Unified simulation and benchmark for garment manipulation tasks β€” folding, hanging, dressing. πŸ“„ πŸ“Š ⭐ - [ClothesNet: An Information-Rich 3D Garment Model Repository with Simulated Clothes Environment](https://arxiv.org/abs/2308.09987) - Large-scale 3D garment dataset with rich annotations for manipulation research. πŸ“„ πŸ“Š - [RAPID: Rapid Adaptation of Particle Dynamics for Generalized Deformable Object Mobile Manipulation](https://arxiv.org/abs/2603.18246) - Rapid adaptation of particle dynamics for generalized deformable object mobile manipulation. πŸ“„ ⭐ - [Benchmarking the Sim-to-Real Gap in Cloth Manipulation](https://arxiv.org/abs/2310.09543) - Systematic evaluation of the sim-to-real gap in cloth manipulation across different simulators. πŸ“„ πŸ“Š - [SoftMimicGen: A Data Generation System for Scalable Robot Learning in Deformable Object Manipulation](https://arxiv.org/abs/2603.25725) - Extends MimicGen to deformable object manipulation, enabling scalable robot demonstration data generation. πŸ“„ ⭐ - [DeformGen: Dynamics-Based Topology Augmentation for Deformable Manipulation Policy Learning](https://arxiv.org/abs/2606.25939) - Dynamics-based topology augmentation for deformable object manipulation policy learning, built on PhysTwin and Real2Render2Real. πŸ“„ ## Sim-to-Real Transfer - [Learning to Manipulate Deformable Objects in the Real World via Sim2Real](https://arxiv.org/abs/2512.11070) - End-to-end sim2real pipeline for deformable object manipulation with domain randomization. πŸ“„ ## Benchmarks & Evaluation - [SoftGym: Benchmarking Deep Reinforcement Learning for Deformable Object Manipulation](https://arxiv.org/abs/2011.07215) - Benchmark suite for deformable object manipulation (cloth, rope, fluid) with standardized RL environments. πŸ“„ πŸ“Š ⭐ ## 3D Asset Generation for Simulation - [PhysX-3D: Physical-Grounded 3D Asset Generation](https://arxiv.org/abs/2507.12465) - End-to-end paradigm for physical-grounded 3D asset generation with PhysXNet dataset. πŸ“„ ⭐ - [PhysX-Anything: Simulation-Ready Physical 3D Assets from Single Image](https://arxiv.org/abs/2511.13468) - Generate simulation-ready 3D assets with physical properties from a single image. πŸ“„ - [PhysX-Omni: Unified Simulation-Ready Physical 3D Generation for Rigid, Deformable, and Articulated Objects](https://arxiv.org/abs/2605.21572) - First to include deformable objects in PhysX generation scope, unifying rigid/deformable/articulated. πŸ“„ ⭐ - [DiffGI: Differentiable Geometry Images](https://arxiv.org/abs/2607.13365) - 3D garment generation via differentiable geometry images (no integrated physics yet). πŸ“„ - [PhysTwin: Physics-Informed Reconstruction and Simulation of Deformable Objects from Videos](https://arxiv.org/abs/2503.17973) - Physics-informed reconstruction and simulation of deformable objects from video; a precursor to Real2Render2Real and DeformGen. πŸ“„ ⭐ - [Gaussian Garments: Reconstructing Simulation-Ready Clothing with Photorealistic Appearance from Multi-View Video](https://arxiv.org/abs/2409.08189) - Reconstructs simulation-ready, photorealistic standalone garment assets from multi-view video. πŸ“„ ## Foundational Simulation Techniques - [XRTailor (OpenXRLab)](https://github.com/openxrlab/xrtailor) - GPU-accelerated cloth simulation engine for large-scale data generation. πŸ”§ - [Position Based Dynamics (PBD)](https://matthias-research.github.io/pages/tenMinutePhysics/) - Classic introductory resource for Position Based Dynamics simulation. πŸ”§ - [ThinShellLab: Thin-Shell Object Manipulations With Differentiable Physics Simulations](https://arxiv.org/abs/2404.00451) - Fully differentiable simulation platform for thin-shell materials (paper, cloth) with varying bending stiffness. πŸ“„ πŸ”§ - [Second-Order FEM for Deformable Surfaces](https://dl.acm.org/doi/10.1145/3592430) - High-order finite element method for improving accuracy in cloth / thin-shell simulation. πŸ“„ --- ## Legend | Tag | Meaning | | --- | ------------------------- | | πŸ“„ | Paper | | πŸ”§ | Tool / Framework / Engine | | πŸ“Š | Benchmark / Dataset | | ⭐ | Recommended / Important | --- ## Contributing Contributions welcome! Please read the [contribution guidelines](CONTRIBUTING.md) first. **Quick way to add a paper:** 1. Edit `papers.yml` β€” add your entry under the appropriate section 2. Run `python generate.py` 3. Submit a Pull Request