Neural Field Movement Primitives for Joint Modelling of Scenes and Motions

TL;DR

This paper introduces a neural field-based Learning from Demonstration method that efficiently learns scene and motion representations, enabling accurate, versatile motion generation for novel scenes without extensive data or handcrafted parameters.

Contribution

It proposes a shared embedding approach for joint scene and motion modeling using neural fields, improving data efficiency and generalization in motion trajectory generation.

Findings

Outperforms baseline methods in accuracy and generalization

Successfully models multi-valued and 6D trajectories

Robust to distractor objects at inference time

Abstract

This paper presents a novel Learning from Demonstration (LfD) method that uses neural fields to learn new skills efficiently and accurately. It achieves this by utilizing a shared embedding to learn both scene and motion representations in a generative way. Our method smoothly maps each expert demonstration to a scene-motion embedding and learns to model them without requiring hand-crafted task parameters or large datasets. It achieves data efficiency by enforcing scene and motion generation to be smooth with respect to changes in the embedding space. At inference time, our method can retrieve scene-motion embeddings using test time optimization, and generate precise motion trajectories for novel scenes. The proposed method is versatile and can employ images, 3D shapes, and any other scene representations that can be modeled using neural fields. Additionally, it can generate both end-effector positions and joint angle-based trajectories. Our method is evaluated on tasks that require accurate motion trajectory generation, where the underlying task parametrization is based on object positions and geometric scene changes. Experimental results demonstrate that the proposed method outperforms the baseline approaches and generalizes to novel scenes. Furthermore, in real-world experiments, we show that our method can successfully model multi-valued trajectories, it is robust to the distractor objects introduced at inference time, and it can generate 6D motions.