Skip to content

Latest commit

 

History

History
25 lines (25 loc) · 3.15 KB

appendix.md

File metadata and controls

25 lines (25 loc) · 3.15 KB
  • bullet3: We refer to panda and xarm robotic arms and the xarm gripper,the underlying simulation is driven by the pybullet engine.
  • PartNet Mobility dataset: We use articulated urdf objects from PartNet Mobility dataset.
  • pybullet-URDF-models: We use rigid urdf objects from pybullet-URDF-models.
  • YCB dataset: We use rigid 3D models from YCB dataset.
  • urdf_files_dataset: We refer to ur5e arm from urdf_files_dataset.
  • flexiv_rdk: We refer to flexiv arm from flexiv_rdk.
  • mycobot_ros: We refer to elephant arm from mycobot_ros.
  • ugv_gazebo_sim: We refer to ranger mini base from ugv_gazebo_sim.
  • dh_ag95_gripper_ros2: We refer to ag-95 gripper from dh_ag95_gripper_ros2.
  • robotiq_arg85_description: We refer to robotiq 2f-85 gripper from robotiq_arg85_description.
  • Wiki-GRx-Models: We refer to humanoid model from Wiki-GRx-Models.
  • ompl: We utilize the algorithms provided by OMPL (Open Motion Planning Library) for motion planning.
  • anygrasp_sdk: We integrate the AnyGrasp algorithm for grasp pose estimation.
  • lang-segment-anything: We integrate the Lang SAM algorithm for object detection.
  • pybullet-blender-recorder: We use pyBulletSimRecorder and pyBulletSimImport from pybullet-blender-recorder to render the simulation scene better.
  • FFmpeg: We use FFmpeg for real-time video recording.
  • isort: We used the isort tool to standardize library imports.
  • black: We use the clack tool to standardize code style.
  • pre-commit: We use pre-commit to automatically run code checks before committing changes.
  • kitchen-worlds: We refer to the scene generation method and scene data of kitchen_world.
  • pybullet_ur5_robotiq: We refer to the gripper model and control of pybullet_ur5_robotiq.
  • urdformer: We integrate urdformer submodule to generate 3D urdf models from 2D images.
  • acdc: We integrate digital cousins submodule to create 3D scene from 2D images.
  • pddlstream: We integrate pddlstream submodule to task planing.
  • LLM+P: We integrate LLM+P submodules to utilize LLM to guide task planning.