HanMoVLM: Large Vision-Language Models for Professional Artistic Painting Evaluation

TL;DR

HanMoVLM is a specialized vision-language model designed to evaluate Chinese artworks professionally, combining expert reasoning, a new dataset, and a reward-based refinement to align with human expert standards.

Contribution

The paper introduces HanMoVLM, a novel model with Chain-of-Thought reasoning and a new dataset, enabling professional-level evaluation of Chinese paintings, bridging the gap between general VLMs and expert judgment.

Findings

HanMoVLM achieves high consistency with professional experts.

The model improves Chinese painting generation quality.

It effectively guides image generation through expert-level verification.

Abstract

While Large Vision-Language Models (VLMs) demonstrate impressive general visual capabilities, they remain artistically blind and unable to offer professional evaluation of artworks within specific artistic domains like human experts. To bridge this gap, we transform VLMs into experts capable of professional-grade painting evaluation in the Chinese Artistic Domain, which is more abstract and demands extensive artistic training for evaluation. We introduce HanMo-Bench, a new dataset that features authentic auction-grade masterpieces and AI-generated works, grounded in real-world market valuations. To realize the rigorous judgment, we propose the HanMoVLM and construct a Chain-of-Thought (CoT) validated by experts. This CoT guides the model to perform expert-level reasoning: from content identification and Region of Interest (RoI) localization to professional evaluation, guided by both theme-specific evaluation and typical three-tier evaluation in Chinese paintings. Furthermore, we design a reward function to refine the reasoning process of the HanMoVLM to improve the accuracy. We demonstrate that HanMoVLM can serve as a critical backbone for Test-time Scaling in image generation. By acting as a high-quality verifier, HanMoVLM enables generative models to select the most artistically superior outputs from multiple candidates. Experimental results and human studies confirm that the proposed HanMoVLM effectively bridges the gap, achieving a high consistency with professional experts and significantly improving the quality of Chinese Painting generation.