A Motion Taxonomy for Manipulation Embedding
This work addresses the need for structured motion representation in robotics and AI, but it is incremental as it builds on existing datasets and methods without major breakthroughs.
The paper tackles the problem of representing motions for manipulation by introducing a motion taxonomy that encodes mechanical properties as binary strings called motion codes, and shows that these codes maintain distances that closely match real manipulation data compared to Word2Vec vectors.
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.