mirror of
https://github.com/wassname/pyrobolearn.git
synced 2026-09-21 13:20:37 +08:00
40 lines
1.2 KiB
Python
40 lines
1.2 KiB
Python
# This file defines the `PoseGenerator` class which generates possible or plausible poses for a robot.
|
|
# A pose is defined as:
|
|
# - the link positions/orientations, and the base position/orientation.
|
|
# - the joint positions, and the base position/orientation.
|
|
|
|
|
|
class PoseGenerator(object):
|
|
|
|
def __init__(self, robot):
|
|
self.robot = robot
|
|
|
|
def generate_uniform_random_pose(self, jnts=None):
|
|
pass
|
|
|
|
def generate_gaussian_random_pose(self, jnts=None):
|
|
"""
|
|
Put a gaussian distribution with the mean sets to jnt initial configuration, and the 2 times
|
|
the standard deviation sets to ...
|
|
:param jnts:
|
|
:return:
|
|
"""
|
|
pass
|
|
|
|
def generate_plausible_pose(self, model, jnts=None):
|
|
"""
|
|
Given a trained learning model (for instance a GAN or VAE), it generates a plausible pose of the robot.
|
|
:param model: learning model
|
|
:param jnts:
|
|
:return:
|
|
"""
|
|
pass
|
|
|
|
def generate_random_pose(self, generator, jnts=None):
|
|
"""
|
|
Based on the given distribution generator, it generates a pose.
|
|
:param generator:
|
|
:param jnts:
|
|
:return:
|
|
"""
|
|
pass |