CARTIER: Cartographic lAnguage Reasoning Targeted at Instruction Execution for Robots

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Rivkin, Dmitriy, Kakodkar, Nikhil, Hogan, Francois, Baghi, Bobak H., Dudek, Gregory
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909090191507456
author Rivkin, Dmitriy
Kakodkar, Nikhil
Hogan, Francois
Baghi, Bobak H.
Dudek, Gregory
author_facet Rivkin, Dmitriy
Kakodkar, Nikhil
Hogan, Francois
Baghi, Bobak H.
Dudek, Gregory
contents This work explores the capacity of large language models (LLMs) to address problems at the intersection of spatial planning and natural language interfaces for navigation. We focus on following complex instructions that are more akin to natural conversation than traditional explicit procedural directives typically seen in robotics. Unlike most prior work where navigation directives are provided as simple imperative commands (e.g., "go to the fridge"), we examine implicit directives obtained through conversational interactions.We leverage the 3D simulator AI2Thor to create household query scenarios at scale, and augment it by adding complex language queries for 40 object types. We demonstrate that a robot using our method CARTIER (Cartographic lAnguage Reasoning Targeted at Instruction Execution for Robots) can parse descriptive language queries up to 42% more reliably than existing LLM-enabled methods by exploiting the ability of LLMs to interpret the user interaction in the context of the objects in the scenario.
format Preprint
id arxiv_https___arxiv_org_abs_2307_11865
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle CARTIER: Cartographic lAnguage Reasoning Targeted at Instruction Execution for Robots
Rivkin, Dmitriy
Kakodkar, Nikhil
Hogan, Francois
Baghi, Bobak H.
Dudek, Gregory
Robotics
Computation and Language
This work explores the capacity of large language models (LLMs) to address problems at the intersection of spatial planning and natural language interfaces for navigation. We focus on following complex instructions that are more akin to natural conversation than traditional explicit procedural directives typically seen in robotics. Unlike most prior work where navigation directives are provided as simple imperative commands (e.g., "go to the fridge"), we examine implicit directives obtained through conversational interactions.We leverage the 3D simulator AI2Thor to create household query scenarios at scale, and augment it by adding complex language queries for 40 object types. We demonstrate that a robot using our method CARTIER (Cartographic lAnguage Reasoning Targeted at Instruction Execution for Robots) can parse descriptive language queries up to 42% more reliably than existing LLM-enabled methods by exploiting the ability of LLMs to interpret the user interaction in the context of the objects in the scenario.
title CARTIER: Cartographic lAnguage Reasoning Targeted at Instruction Execution for Robots
topic Robotics
Computation and Language
url https://arxiv.org/abs/2307.11865