ReasonNav: Human-Inspired Global Map Reasoning for Zero-Shot Embodied Navigation

AO Yuzhuo, Anbang Wang, Yu-Wing Tai, Chi-Keung Tang

Keywords: Large Language Model, Multimodalitiy, Embodied navigation

TL;DR: In this paper, we introduces ReasonNav, a novel framework for zero-shot embodied navigation, leveraging the strengths of MLLMs for high-level hierarchical global reasoning, while relegating low-level control to robust, deterministic planners.

Abstract: Embodied agents often struggle with efficient navigation because they rely solely on partial egocentric observations, which restrict global foresight and lead to inefficient exploration. In contrast, humans plan using maps: we reason globally first, then act locally. We introduce ReasonNav, a human-inspired framework that operationalizes this reason-then-act paradigm by coupling Multimodal Large Language Models (MLLMs) with deterministic planners. ReasonNav converts a top-down map into a discrete reasoning space by room segmentation and candidate target nodes sampling. An MLLM is then queried in a multi-stage process to identify the candidate most consistent with the instruction (object, image, or text goal), effectively leveraging the model’s semantic reasoning ability while sidestepping its weakness in continuous coordinate prediction. The selected waypoint is grounded into executable trajectories using a deterministic action planner over an online-built occupancy map, while pretrained object detectors and segmenters ensure robust recognition at the goal. This yields a unified zero-shot navigation framework that requires no MLLM fine-tuning, circumvents the brittleness of RL-based policies and scales naturally with foundation model improvements. Across three navigation tasks, ReasonNav consistently outperforms prior methods that demand extensive training or heavy scene modeling, offering a scalable, interpretable, and globally grounded solution to embodied navigation.

Primary Area: applications to computer vision, audio, language, and other modalities

Submission Number: 6839