root directory
├── assets
│ └── // folders for different YCB and real scan object
├── data
│ └── // folders for data after optimization
├── gpis_states
│ └── // state data for restoring and visualizing gaussian process implicit surface
├── pybullet_robot
│ ├── // fingertip impedance and null space controller for leap hand and allegro hand
| └── src
| └── // source code for robot and controller definition
├── thirdparty
│ ├── // dependent third-party package
| └── TorchSDF
| └── // Compute SDF of mesh in pytorch
| └── differentiable-robot-model
| └── // Differentiable forward kinematics model
├── 3dplot.py // Visualizing GPIS intersection and its uncertainty
├── concatenate_pcd.py // Merging generated point cloud with observed pointcloud
├── gpis.py // Definition for Gaussian process implicit surface
├── optimize_pregrasp.py // Running probabilistic pregrasp optimization
├── rigidBodySento.py // Some useful tools for creating and handling Pybullet rigid objects
├── train_gpis_completion.py // Training GPIS from observed + completed point clouds
└── verify_pregrasp.py // verifying pregrasp by simulation
Ericcsr/ComplianceDex
Compliance dexterity project basedo Isaac Gym Reinforcement Learning Environments
Python