Houdini 20.5 Nodes Geometry nodes

ML Pose Generate geometry node

Generate a set of random poses.

On this page

Overview

ML Pose Generate is a utility that allows you to generate a set of random poses. Each pose is a copy of the input which a specified group of joints has a randomized transform. The random poses are stored inside packed primitives.

See Machine Learning documentation for more general information.

Context

Each random pose is recognized by the ML tools as an unlabeled example, see ML Example. Given a set of random poses, a set of labeled examples may be formed by performing a computation for each pose, for example deforming a skin using simulation, and using ML Example afterwards.

Using ML Pose Generate to take existing joint limits and generate random poses in a Gaussian distribution. Example from ML Groom Deformer Tutorial

Parameters

Random Seed

Random seed used to generate all random numbers in each of the samples.

Sample Count

The number of samples that are generated.

Joint Group

This group indicates the specific joints that are to be serialized.

Inputs

Pose Prototype

Each generated sample is a copy of this input, with the joint transforms randomized.

Outputs

Unlabeled Pose Examples

A geometry consisting entirely of packed primitives, each contains a copy of Prototype with the designated attribute randomized.

Geometry nodes