See2Refine: Vision-Language Feedback Improves LLM-Based eHMI Action Designers

Automated vehicles lack natural communication channels with other road users, making external Human-Machine Interfaces (eHMIs) essential to convey intent and maintain trust in shared environments. However, most eHMI studies rely on developer-crafted messageaction pairs, which are difficult to adapt to diverse and dynamic traffic contexts. A promising alternative is to use Large Language Models (LLMs) as action designers that generate context-conditioned eHMI actions, yet such designers lack perceptual verification and typically depend on fixed prompts or costly humanannotated feedback for improvement. We present SEE2REFINE, a human-free, closedloop framework that uses vision-language models (VLMs) for perceptual evaluation as automated visual feedback to improve an LLMbased eHMI action designer. Given a driving context and a candidate eHMI action, the VLM evaluates the perceived appropriateness of the action, and this feedback is used to iteratively revise the designer's output, enabling systematic refinement without human supervision. We evaluate our framework across three eHMI modalities (lightbar, eyes, and arm) and multiple LLM model sizes. Across settings, our framework consistently outperforms prompt-only LLM designers and manually specified baselines in both VLM-based metrics and human-subject evaluations. The results further indicate that the improvements are generalized across modalities and that VLM evaluations are reasonably aligned with human preferences in our controlled settings, supporting the robustness and effectiveness of SEE2REFINE for scalable action design. 1