Hierarchical Cross-Modal Agent for Robotics Vision-and-Language Navigation

TL;DR

This paper introduces Robo-VLN, a more realistic continuous environment for vision-and-language navigation, and proposes a hierarchical agent that significantly improves navigation performance over existing methods.

Contribution

It presents Robo-VLN, a new continuous 3D environment for VLN, and develops a hierarchical agent with layered decision-making that outperforms prior approaches.

Findings

Hierarchical agent outperforms baselines in Robo-VLN

Layered decision-making improves navigation accuracy

Modular training enhances policy effectiveness

Abstract

Deep Learning has revolutionized our ability to solve complex problems such as Vision-and-Language Navigation (VLN). This task requires the agent to navigate to a goal purely based on visual sensory inputs given natural language instructions. However, prior works formulate the problem as a navigation graph with a discrete action space. In this work, we lift the agent off the navigation graph and propose a more complex VLN setting in continuous 3D reconstructed environments. Our proposed setting, Robo-VLN, more closely mimics the challenges of real world navigation. Robo-VLN tasks have longer trajectory lengths, continuous action spaces, and challenges such as obstacles. We provide a suite of baselines inspired by state-of-the-art works in discrete VLN and show that they are less effective at this task. We further propose that decomposing the task into specialized high- and low-level policies can more effectively tackle this task. With extensive experiments, we show that by using layered decision making, modularized training, and decoupling reasoning and imitation, our proposed Hierarchical Cross-Modal (HCM) agent outperforms existing baselines in all key metrics and sets a new benchmark for Robo-VLN.