用结构化表示提升图表转代码的准确性
Chart Specification: Structural Representations for Incentivizing VLM Reasoning in Chart-to-Code Generation
- 引入图表规格化中间表示，避免表面文本模仿
- 3000样本下性能领先基线61.7%，4000样本达新纪录
- 适合需要高精度图表代码生成的研究者与开发者
视觉语言模型（VLM）在从图表图像生成绘图代码方面展现出潜力，但保持结构一致性仍具挑战。现有方法多依赖监督微调，导致模型倾向于表面级标记模仿而非忠实建模底层图表结构，常产生幻觉或语义不一致输出。本文提出图表规格化（Chart Specification），一种结构化的中间表示，将训练目标从文本模仿转向语义基础的监督。该方法通过过滤语法噪声构建结构平衡的训练集，并支持细粒度可验证的Spec-Align奖励机制，使强化学习能有效约束绘图逻辑的一致性。在三个公开基准上的实验表明，本方法持续优于先前方法。仅使用3000个训练样本即实现优异数据效率，在复杂基准上超越领先基线最高达61.7%；扩展至4000样本后，在所有评估指标上均达到新最优结果。结果表明，精确的结构化监督是实现高保真图表转代码的有效路径。代码与数据集见：https://github.com/Mighten/chart-specification-paper
原文摘要 · Abstract (English)
Vision-Language Models (VLMs) have shown promise in generating plotting code from chart images, yet achieving structural fidelity remains challenging. Existing approaches largely rely on supervised fine-tuning, encouraging surface-level token imitation rather than faithful modeling of underlying chart structure, which often leads to hallucinated or semantically inconsistent outputs. We propose Chart Specification, a structured intermediate representation that shifts training from text imitation to semantically grounded supervision. Chart Specification filters syntactic noise to construct a structurally balanced training set and supports a Spec-Align Reward that provides fine-grained, verifiable feedback on structural correctness, enabling reinforcement learning to enforce consistent plotting logic. Experiments on three public benchmarks show that our method consistently outperforms prior approaches. With only 3K training samples, we achieve strong data efficiency, surpassing leading baselines by up to 61.7% on complex benchmarks, and scaling to 4K samples establishes new state-of-the-art results across all evaluated metrics. Overall, our results demonstrate that precise structural supervision offers an efficient pathway to high-fidelity chart-to-code generation. Code and dataset are available at: https://github.com/Mighten/chart-specification-paper
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。