MAction-SocialNav: Multi-Action Socially Compliant Navigation via Reasoning-enhanced Prompt Tuning

TL;DR

This paper introduces MAction-SocialNav, a vision-language model that generates multiple socially compliant navigation actions, improving reasoning and safety in ambiguous social scenarios for robots.

Contribution

The work presents a novel multi-action socially compliant navigation model with reasoning-enhanced prompt tuning and a new dataset for diverse social navigation scenarios.

Findings

Outperforms GPT-4o and Claude in decision quality and safety alignment.

Achieves real-time performance at over 3x faster speed.

Demonstrates strong social reasoning and efficiency in complex environments.

Abstract

Socially compliant navigation requires robots to move safely and appropriately in human-centered environments by respecting social norms. However, social norms are often ambiguous, and in a single scenario, multiple actions may be equally acceptable. Most existing methods simplify this problem by assuming a single correct action, which limits their ability to handle real-world social uncertainty. In this work, we propose MAction-SocialNav, an efficient vision language model for socially compliant navigation that explicitly addresses action ambiguity, enabling generating multiple plausible actions within one scenario. To enhance the model's reasoning capability, we introduce a novel meta-cognitive prompt (MCP) method. Furthermore, to evaluate the proposed method, we curate a multi-action socially compliant navigation dataset that accounts for diverse conditions, including crowd density, indoor and outdoor environments, and dual human annotations. The dataset contains 789 samples, each with three-turn conversation, split into 710 training samples and 79 test samples through random selection. We also design five evaluation metrics to assess high-level decision precision, safety, and diversity. Extensive experiments demonstrate that the proposed MAction-SocialNav achieves strong social reasoning performance while maintaining high efficiency, highlighting its potential for real-world human robot navigation. Compared with zero-shot GPT-4o and Claude, our model achieves substantially higher decision quality (APG: 0.595 vs. 0.000/0.025) and safety alignment (ER: 0.264 vs. 0.642/0.668), while maintaining real-time efficiency (1.524 FPS, over 3x faster).