Masaki Murooka - Software

Baseline for Robot Learning

RoboManipBaselines

A software framework integrating various imitation learning methods and benchmark environments for robotic manipulation.

Baseline Humanoid Controllers

This section presents open-source humanoid controllers developed based on the robot control framework mc_rtc. All of the following controllers are implemented in a unified architecture and enable general baseline motions of humanoids.

BaselineWalkingController

Humanoid walking controller with various baseline methods.

LocomanipController

Humanoid loco-manipulation controller.

MultiContactController

Humanoid multi-contact motion controller.

HumanRetargetingController

Controller for retargeting the motion from human to humanoid robot.

Baseline Humanoid Planner

BaselineFootstepPlanner

Humanoid footstep planner based on baseline methods with graph search.

General-purpose Controller

NMPC

Non-linear model predictive control (NMPC) library.

Simulation Utilities

MujocoRosUtils

ROS-based MuJoCo utilities.

MujocoTactileSensorPlugin

Plugin to simulate tactile sensors in MuJoCo.

CnoidRosUtils

ROS-based Choreonoid utilities.

Visualization Utilities

RobotArrayVisualization

ROS-based visualization library for robot arrays.

MocapRosUtils

ROS-based utilities for motion capture.

Simulation Challenge

Participated in the Humanoid Virtual Athletics Challenge 2021 as Team AIST-JRL-MM.