Learning and Inferring Movement with Deep Generative Model

TL;DR

This paper introduces a probabilistic deep generative model for learning and inferring complex movements in robotics, effectively handling high-dimensional data and varied environments, with demonstrated advantages over traditional methods.

Contribution

It presents a novel probabilistic framework using deep generative models for movement learning and inference, incorporating task and context information for improved robotic path planning.

Findings

Outperforms baseline methods with limited training data

Successfully incorporates task descriptors and context for long-term planning

Demonstrates effectiveness in robotic approaching path planning

Abstract

Learning and inference movement is a very challenging problem due to its high dimensionality and dependency to varied environments or tasks. In this paper, we propose an effective probabilistic method for learning and inference of basic movements. The motion planning problem is formulated as learning on a directed graphic model and deep generative model is used to perform learning and inference from demonstrations. An important characteristic of this method is that it flexibly incorporates the task descriptors and context information for long-term planning and it can be combined with dynamic systems for robot control. The experimental validations on robotic approaching path planning tasks show the advantages over the base methods with limited training data.