The Robot's Inner Critic: Self-Refinement of Social Behaviors through VLM-based Replanning

TL;DR

This paper introduces CRISP, an autonomous framework enabling robots to critique and refine their social behaviors using a Vision-Language Model, significantly improving naturalness and appropriateness across various robot platforms.

Contribution

CRISP is a novel, general framework that allows robots to self-critique and replan social behaviors autonomously using VLMs, reducing human intervention and enhancing cross-platform applicability.

Findings

Higher preference ratings in user studies

Improved situational appropriateness of behaviors

Effective self-refinement across multiple robot types

Abstract

Conventional robot social behavior generation has been limited in flexibility and autonomy, relying on predefined motions or human feedback. This study proposes CRISP (Critique-and-Replan for Interactive Social Presence), an autonomous framework where a robot critiques and replans its own actions by leveraging a Vision-Language Model (VLM) as a `human-like social critic.' CRISP integrates (1) extraction of movable joints and constraints by analyzing the robot's description file (e.g., MJCF), (2) generation of step-by-step behavior plans based on situational context, (3) generation of low-level joint control code by referencing visual information (joint range-of-motion visualizations), (4) VLM-based evaluation of social appropriateness and naturalness, including pinpointing erroneous steps, and (5) iterative refinement of behaviors through reward-based search. This approach is not tied to a specific robot API; it can generate subtly different, human-like motions on various platforms using only the robot's structure file. In a user study involving five different robot types and 20 scenarios, including mobile manipulators and humanoids, our proposed method achieved significantly higher preference and situational appropriateness ratings compared to previous methods. This research presents a general framework that minimizes human intervention while expanding the robot's autonomous interaction capabilities and cross-platform applicability. Detailed result videos and supplementary information regarding this work are available at: https://limjiyu99.github.io/inner-critic/