Timezone: »

 
Skill Acquisition by Instruction Augmentation on Offline Datasets
Ted Xiao · Harris Chan · Pierre Sermanet · Ayzaan Wahid · Anthony Brohan · Karol Hausman · Sergey Levine · Jonathan Tompson

@
in Foundation Models for Decision Making »
Event URL: https://openreview.net/forum?id=uztKyepCdd »

In recent years, much progress has been made in learning robotic manipulation policies that follow natural language instructions. Commonly, such methods learn from a corpora of robot-language data that was either collected with specific tasks in mind or expensively re-labelled by humans with rich language descriptions in hindsight. Recently, large-scale pretrained vision-language models like CLIP have been applied to robotics in the form of learning representations and planners. Can these pretrained models also be used to cheaply impart internet-scale knowledge onto offline datasets, providing access to skills that were not reflected in ground truth labels? To accomplish this, we introduce Data-driven Instruction Augmentation for Language-conditioned control (DIAL): we utilize semi-supervised language labels leveraging the semantic understanding of CLIP to propagate knowledge onto large datasets of unlabelled demonstration data and then train language-conditioned policies on the augmented datasets. This method enables cheaper acquisition of useful language descriptions compared to expensive human labels, allowing for more efficient label coverage of large-scale datasets. We apply DIAL to a challenging real-world robotic manipulation domain, enabling imitation learning policies to acquire new capabilities and generalize to 80 novel instructions unseen in the original dataset.

Author Information

Ted Xiao (Google Brain)
Harris Chan (University of Toronto, Vector Institute)
Pierre Sermanet (Google Brain)
Ayzaan Wahid (Google)
Anthony Brohan (Google Research)
Karol Hausman (Google Brain)
Sergey Levine (UC Berkeley)
Jonathan Tompson (Google Brain)

More from the Same Authors