ChartE$^{3}$: A Comprehensive Benchmark for End-to-End Chart Editing

TL;DR

ChartE3 tackles end-to-end chart editing by introducing a direct image-to-image benchmark that bypasses intermediate code representations. It constructs a five-stage data pipeline to provide over 800 chart images and 1,200 editing samples across 12 tasks and 10 chart types, evaluated with a hybrid of SSIM, PSNR, CLIP, DINO, LPIPS, and GPT-based Correctness/Consistency judgments. Experiments reveal substantial gaps in current multimodal models, especially for global, data-centric edits, and demonstrate strong alignment between GPT-based and human evaluations. The work establishes a practical framework and metrics for developing structure-aware, faithfully edited charts, guiding future research toward robust end-to-end chart editing systems.

Abstract

Charts are a fundamental visualization format for structured data analysis. Enabling end-to-end chart editing according to user intent is of great practical value, yet remains challenging due to the need for both fine-grained control and global structural consistency. Most existing approaches adopt pipeline-based designs, where natural language or code serves as an intermediate representation, limiting their ability to faithfully execute complex edits. We introduce ChartE$^{3}$, an End-to-End Chart Editing benchmark that directly evaluates models without relying on intermediate natural language programs or code-level supervision. ChartE$^{3}$ focuses on two complementary editing dimensions: local editing, which involves fine-grained appearance changes such as font or color adjustments, and global editing, which requires holistic, data-centric transformations including data filtering and trend line addition. ChartE$^{3}$ contains over 1,200 high-quality samples constructed via a well-designed data pipeline with human curation. Each sample is provided as a triplet of a chart image, its underlying code, and a multimodal editing instruction, enabling evaluation from both objective and subjective perspectives. Extensive benchmarking of state-of-the-art multimodal large language models reveals substantial performance gaps, particularly on global editing tasks, highlighting critical limitations in current end-to-end chart editing capabilities.

Figures (7)

• Figure 1: Overview of the five-stage data construction pipeline for ChartE$^{3}$.
• Figure 2: The distribution of chart types and editing tasks.
• Figure 3: The error analysis of Nano Banana and Qwen-Image-Edit.
• Figure 4: Editing categories: Axis, Conversion, and Color.
• Figure 5: Editing categories: Filtering, Font, and Legend.