Breaking Down and Building Up: Mixture of Skill-Based Vision-and-Language Navigation Agents

TL;DR

SkillNav introduces a modular, skill-based framework for Vision-and-Language Navigation, enhancing interpretability and generalization in complex, unseen environments through structured reasoning and a novel routing mechanism.

Contribution

The paper presents SkillNav, a new modular approach that decomposes navigation into interpretable skills and employs a dynamic VLM-based router for improved generalization.

Findings

Achieves state-of-the-art generalization on GSA-R2R benchmark.

Constructs a synthetic dataset pipeline for skill-specific instruction-trajectory pairs.

Demonstrates competitive results on standard VLN benchmarks.

Abstract

Vision-and-Language Navigation (VLN) poses significant challenges for agents to interpret natural language instructions and navigate complex 3D environments. While recent progress has been driven by large-scale pre-training and data augmentation, current methods still struggle to generalize to unseen scenarios, particularly when complex spatial and temporal reasoning is required. In this work, we propose SkillNav, a modular framework that introduces structured, skill-based reasoning into Transformer-based VLN agents. Our method decomposes navigation into a set of interpretable atomic skills (e.g., Vertical Movement, Area and Region Identification, Stop and Pause), each handled by a specialized agent. To support targeted skill training without manual data annotation, we construct a synthetic dataset pipeline that generates diverse, linguistically natural, skill-specific instruction-trajectory pairs. We then introduce a novel training-free Vision-Language Model (VLM)-based router, which dynamically selects the most suitable agent at each time step by aligning sub-goals with visual observations and historical actions. SkillNav obtains competitive results on commonly used benchmarks and establishes state-of-the-art generalization to the GSA-R2R, a benchmark with novel instruction styles and unseen environments.