EvoChart: A Benchmark and a Self-Training Approach Towards Real-World Chart Understanding

TL;DR

EvoChart introduces a self-training data synthesis method and a comprehensive benchmark to improve and evaluate real-world chart understanding capabilities of visual language models, addressing data scarcity and evaluation gaps.

Contribution

The paper presents EvoChart, a novel self-training approach for synthetic chart data generation and a new benchmark, EvoChart-QA, for assessing model performance in real-world chart comprehension.

Findings

EvoChart significantly improves VLMs' accuracy on real-world chart tasks.

GPT-4o achieves only 49.8% accuracy on the benchmark.

Open-source VLMs' performance is boosted to 54.2% with EvoChart.

Abstract

Chart understanding enables automated data analysis for humans, which requires models to achieve highly accurate visual comprehension. While existing Visual Language Models (VLMs) have shown progress in chart understanding, the lack of high-quality training data and comprehensive evaluation benchmarks hinders VLM chart comprehension. In this paper, we introduce EvoChart, a novel self-training method for generating synthetic chart data to enhance VLMs' capabilities in real-world chart comprehension. We also propose EvoChart-QA, a noval benchmark for measuring models' chart comprehension abilities in real-world scenarios. Specifically, EvoChart is a unique self-training data synthesis approach that simultaneously produces high-quality training corpus and a high-performance chart understanding model. EvoChart-QA consists of 650 distinct real-world charts collected from 140 different websites and 1,250 expert-curated questions that focus on chart understanding. Experimental results on various open-source and proprietary VLMs tested on EvoChart-QA demonstrate that even the best proprietary model, GPT-4o, achieves only 49.8% accuracy. Moreover, the EvoChart method significantly boosts the performance of open-source VLMs on real-world chart understanding tasks, achieving 54.2% accuracy on EvoChart-QA.