Search papers, labs, and topics across Lattice.
This paper introduces VizAnchor, a novel framework designed to decode manipulation intent in data visualizations through dual-anchor reasoning. By constructing both semantic and spatial anchors, VizAnchor effectively recovers authentic chart information and localizes tampered regions, while three specialized agents analyze and explain the manipulation. The evaluation demonstrates that VizAnchor not only accurately identifies manipulations but also provides coherent explanations of the misleading intents behind them, marking a significant advancement in understanding visualization integrity.
VizAnchor reveals the hidden motives behind data visualization manipulations, offering clarity in a landscape rife with misleading interpretations.
Data visualizations are widely used for communicating information, but they are also vulnerable to intentional manipulations that induce misleading interpretations. Existing methods focus on locating tampered regions or recovering hidden information, without explaining how the visualization has been manipulated or why the resulting changes may mislead viewers. We propose \textbf{VizAnchor}, a framework for visualization manipulation understanding through dual-anchor evidence construction and VLM-based reasoning. In the first stage, VizAnchor constructs a semantic anchor to recover authentic chart information and a spatial anchor to localize tampered regions. In the second stage, three specialized agents decode the manipulation. The misleader grounding agent analyzes a four-panel visual prompt to predict the misleader information. The chart narrative reconstruction agent takes the original and tampered charts as inputs and reconstructs their respective visual narratives. Finally, the intent inferring agent integrates the visual evidence and misleader information to infer the misleading intent. We further construct a dataset for tampering localization and a dataset for misleading intent inferring. Evaluation shows that VizAnchor accurately localizes manipulations and produces faithful explanations of their manipulation, misleaders, and misleading intents.
