Goal Representations for Instruction Following: A Semi-Supervised Language Interface to Control
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2023
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910776253480960 |
|---|---|
| author | Myers, Vivek He, Andre Fang, Kuan Walke, Homer Hansen-Estruch, Philippe Cheng, Ching-An Jalobeanu, Mihai Kolobov, Andrey Dragan, Anca Levine, Sergey |
| author_facet | Myers, Vivek He, Andre Fang, Kuan Walke, Homer Hansen-Estruch, Philippe Cheng, Ching-An Jalobeanu, Mihai Kolobov, Andrey Dragan, Anca Levine, Sergey |
| contents | Our goal is for robots to follow natural language instructions like "put the towel next to the microwave." But getting large amounts of labeled data, i.e. data that contains demonstrations of tasks labeled with the language instruction, is prohibitive. In contrast, obtaining policies that respond to image goals is much easier, because any autonomous trial or demonstration can be labeled in hindsight with its final state as the goal. In this work, we contribute a method that taps into joint image- and goal- conditioned policies with language using only a small amount of language data. Prior work has made progress on this using vision-language models or by jointly training language-goal-conditioned policies, but so far neither method has scaled effectively to real-world robot tasks without significant human annotation. Our method achieves robust performance in the real world by learning an embedding from the labeled data that aligns language not to the goal image, but rather to the desired change between the start and goal images that the instruction corresponds to. We then train a policy on this embedding: the policy benefits from all the unlabeled data, but the aligned embedding provides an interface for language to steer the policy. We show instruction following across a variety of manipulation tasks in different scenes, with generalization to language instructions outside of the labeled data. Videos and code for our approach can be found on our website: https://rail-berkeley.github.io/grif/ . |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2307_00117 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Goal Representations for Instruction Following: A Semi-Supervised Language Interface to Control Myers, Vivek He, Andre Fang, Kuan Walke, Homer Hansen-Estruch, Philippe Cheng, Ching-An Jalobeanu, Mihai Kolobov, Andrey Dragan, Anca Levine, Sergey Robotics Machine Learning Our goal is for robots to follow natural language instructions like "put the towel next to the microwave." But getting large amounts of labeled data, i.e. data that contains demonstrations of tasks labeled with the language instruction, is prohibitive. In contrast, obtaining policies that respond to image goals is much easier, because any autonomous trial or demonstration can be labeled in hindsight with its final state as the goal. In this work, we contribute a method that taps into joint image- and goal- conditioned policies with language using only a small amount of language data. Prior work has made progress on this using vision-language models or by jointly training language-goal-conditioned policies, but so far neither method has scaled effectively to real-world robot tasks without significant human annotation. Our method achieves robust performance in the real world by learning an embedding from the labeled data that aligns language not to the goal image, but rather to the desired change between the start and goal images that the instruction corresponds to. We then train a policy on this embedding: the policy benefits from all the unlabeled data, but the aligned embedding provides an interface for language to steer the policy. We show instruction following across a variety of manipulation tasks in different scenes, with generalization to language instructions outside of the labeled data. Videos and code for our approach can be found on our website: https://rail-berkeley.github.io/grif/ . |
| title | Goal Representations for Instruction Following: A Semi-Supervised Language Interface to Control |
| topic | Robotics Machine Learning |
| url | https://arxiv.org/abs/2307.00117 |