Search papers, labs, and topics across Lattice.
This paper introduces OmniDiagram, a unified framework for generating diverse diagram types from code, addressing the limitations of existing systems in task formulation and language support. To improve code-to-visual alignment, they propose Visual Interrogation Verifies All (\textsc{Viva}), a novel visual feedback strategy that uses generative visual inquiries to reward the visual structure of rendered diagrams, eliminating the need for ground truth code. Experiments on their new large-scale M3$^2$Diagram dataset (196k instances) demonstrate that OmniDiagram, trained with supervised fine-tuning and \textsc{Viva}-based RL, achieves state-of-the-art results across diagram code generation benchmarks.
Forget hand-annotated code: OmniDiagram uses a self-evolving visual interrogation technique to train diagram-generating AI, achieving state-of-the-art results.
The paradigm of programmable diagram generation is evolving rapidly, playing a crucial role in structured visualization. However, most existing studies are confined to a narrow range of task formulations and language support, constraining their applicability to diverse diagram types. In this work, we propose OmniDiagram, a unified framework that incorporates diverse diagram code languages and task definitions. To address the challenge of aligning code logic with visual fidelity in Reinforcement Learning (RL), we introduce a novel visual feedback strategy named Visual Interrogation Verifies All (\textsc{Viva}). Unlike brittle syntax-based rules or pixel-level matching, \textsc{Viva} rewards the visual structure of rendered diagrams through a generative approach. Specifically, \textsc{Viva} actively generates targeted visual inquiries to scrutinize diagram visual fidelity and provides fine-grained feedback for optimization. This mechanism facilitates a self-evolving training process, effectively obviating the need for manually annotated ground truth code. Furthermore, we construct M3$^2$Diagram, the first large-scale diagram code generation dataset, containing over 196k high-quality instances. Experimental results confirm that the combination of SFT and our \textsc{Viva}-based RL allows OmniDiagram to establish a new state-of-the-art (SOTA) across diagram code generation benchmarks.
