Spatial-Language Attention Policies for Efficient Robot Learning

TL;DR

This paper introduces Spatial-Language Attention Policies (SLAP), a novel approach for mobile manipulation that uses 3D tokens and language-conditioned policies to improve robustness and efficiency in real-world tasks, even with limited data.

Contribution

SLAP is the first method to effectively combine 3D spatial language tokens with multi-task learning for mobile robot manipulation, achieving high success rates with limited data.

Findings

80% success rate in real-world tasks with a single model

47.5% success rate with unseen clutter and configurations

4x improvement over baseline in mobile manipulation

Abstract

Despite great strides in language-guided manipulation, existing work has been constrained to table-top settings. Table-tops allow for perfect and consistent camera angles, properties are that do not hold in mobile manipulation. Task plans that involve moving around the environment must be robust to egocentric views and changes in the plane and angle of grasp. A further challenge is ensuring this is all true while still being able to learn skills efficiently from limited data. We propose Spatial-Language Attention Policies (SLAP) as a solution. SLAP uses three-dimensional tokens as the input representation to train a single multi-task, language-conditioned action prediction policy. Our method shows an 80% success rate in the real world across eight tasks with a single model, and a 47.5% success rate when unseen clutter and unseen object configurations are introduced, even with only a handful of examples per task. This represents an improvement of 30% over prior work (20% given unseen distractors and configurations). We see a 4x improvement over baseline in mobile manipulation setting. In addition, we show how SLAPs robustness allows us to execute Task Plans from open-vocabulary instructions using a large language model for multi-step mobile manipulation. For videos, see the website: https://robotslap.github.io