A Motion Taxonomy for Manipulation Embedding

TL;DR

This paper introduces a motion taxonomy that encodes manipulation motions as binary strings called motion codes, capturing mechanical properties for better analysis and learning.

Contribution

It proposes a novel motion encoding scheme using motion codes that reflect mechanical properties, improving manipulation representation and clustering.

Findings

Motion codes effectively capture mechanical properties of manipulations.

Motion codes maintain realistic distance metrics compared to Word2Vec vectors.

The approach facilitates clustering and analysis of manipulation actions.

Abstract

To represent motions from a mechanical point of view, this paper explores motion embedding using the motion taxonomy. With this taxonomy, manipulations can be described and represented as binary strings called motion codes. Motion codes capture mechanical properties, such as contact type and trajectory, that should be used to define suitable distance metrics between motions or loss functions for deep learning and reinforcement learning. Motion codes can also be used to consolidate aliases or cluster motion types that share similar properties. Using existing data sets as a reference, we discuss how motion codes can be created and assigned to actions that are commonly seen in activities of daily living based on intuition as well as real data. Motion codes are compared to vectors from pre-trained Word2Vec models, and we show that motion codes maintain distances that closely match the reality of manipulation.