SOON: Scenario Oriented Object Navigation with Graph-based Exploration

TL;DR

This paper introduces SOON, a new scenario-oriented object navigation task in 3D environments, with a graph-based exploration method and a large-scale benchmark to improve navigation from arbitrary starting points.

Contribution

It proposes a novel graph-based exploration approach and a new benchmark dataset for scenario-oriented object navigation in 3D environments.

Findings

GBE outperforms state-of-the-art methods on FAO and R2R datasets.

The FAO dataset provides rich semantic descriptions to reduce target ambiguity.

Ablation studies validate the effectiveness of the dataset and method.

Abstract

The ability to navigate like a human towards a language-guided target from anywhere in a 3D embodied environment is one of the 'holy grail' goals of intelligent robots. Most visual navigation benchmarks, however, focus on navigating toward a target from a fixed starting point, guided by an elaborate set of instructions that depicts step-by-step. This approach deviates from real-world problems in which human-only describes what the object and its surrounding look like and asks the robot to start navigation from anywhere. Accordingly, in this paper, we introduce a Scenario Oriented Object Navigation (SOON) task. In this task, an agent is required to navigate from an arbitrary position in a 3D embodied environment to localize a target following a scene description. To give a promising direction to solve this task, we propose a novel graph-based exploration (GBE) method, which models the navigation state as a graph and introduces a novel graph-based exploration approach to learn knowledge from the graph and stabilize training by learning sub-optimal trajectories. We also propose a new large-scale benchmark named From Anywhere to Object (FAO) dataset. To avoid target ambiguity, the descriptions in FAO provide rich semantic scene information includes: object attribute, object relationship, region description, and nearby region description. Our experiments reveal that the proposed GBE outperforms various state-of-the-arts on both FAO and R2R datasets. And the ablation studies on FAO validates the quality of the dataset.