KidsArtBench: Multi-Dimensional Children's Art Evaluation with Attribute-Aware MLLMs

TL;DR

This paper introduces KidsArtBench, a comprehensive dataset of children's artwork with multi-dimensional expert annotations, and proposes an attribute-aware evaluation method for multimodal models to better assess artistic qualities in educational contexts.

Contribution

The paper presents KidsArtBench, a new multi-dimensional benchmark for children's art evaluation, and develops an attribute-specific multi-LoRA approach with RAFT for improved assessment accuracy.

Findings

Increased correlation from 0.468 to 0.653 on evaluation metrics.

Enhanced assessment of perceptual and higher-order attributes.

Established a new testbed for educational AI evaluation.

Abstract

Multimodal Large Language Models (MLLMs) show remarkable progress across many visual-language tasks; however, their capacity to evaluate artistic expression remains limited. Aesthetic concepts are inherently abstract and open-ended, and multimodal artwork annotations are scarce. We introduce KidsArtBench, a new benchmark of over 1k children's artworks (ages 5-15) annotated by 12 expert educators across 9 rubric-aligned dimensions, together with expert comments for feedback. Unlike prior aesthetic datasets that provide single scalar scores on adult imagery, KidsArtBench targets children's artwork and pairs multi-dimensional annotations with comment supervision to enable both ordinal assessment and formative feedback. Building on this resource, we propose an attribute-specific multi-LoRA approach, where each attribute corresponds to a distinct evaluation dimension (e.g., Realism, Imagination) in the scoring rubric, with Regression-Aware Fine-Tuning (RAFT) to align predictions with ordinal scales. On Qwen2.5-VL-7B, our method increases correlation from 0.468 to 0.653, with the largest gains on perceptual dimensions and narrowed gaps on higher-order attributes. These results show that educator-aligned supervision and attribute-aware training yield pedagogically meaningful evaluations and establish a rigorous testbed for sustained progress in educational AI. We release data and code with ethics documentation.