Semantic Document Derendering: SVG Reconstruction via Vision-Language Modeling

TL;DR

This paper presents SliDer, a novel framework using Vision-Language Models to convert raster slide images into editable SVG formats, preserving semantic structure and enabling better document editing.

Contribution

Introduces SliDer, a new method leveraging Vision-Language Models for semantic derendering of slide images into structured SVGs, along with the Slide2SVG dataset for future research.

Findings

SliDer achieves a reconstruction LPIPS of 0.069.

Human evaluators prefer SliDer in 82.9% of cases.

Outperforms zero-shot VLM baseline in semantic reconstruction.

Abstract

Multimedia documents such as slide presentations and posters are designed to be interactive and easy to modify. Yet, they are often distributed in a static raster format, which limits editing and customization. Restoring their editability requires converting these raster images back into structured vector formats. However, existing geometric raster-vectorization methods, which rely on low-level primitives like curves and polygons, fall short at this task. Specifically, when applied to complex documents like slides, they fail to preserve the high-level structure, resulting in a flat collection of shapes where the semantic distinction between image and text elements is lost. To overcome this limitation, we address the problem of semantic document derendering by introducing SliDer, a novel framework that uses Vision-Language Models (VLMs) to derender slide images as compact and editable Scalable Vector Graphic (SVG) representations. SliDer detects and extracts attributes from individual image and text elements in a raster input and organizes them into a coherent SVG format. Crucially, the model iteratively refines its predictions during inference in a process analogous to human design, generating SVG code that more faithfully reconstructs the original raster upon rendering. Furthermore, we introduce Slide2SVG, a novel dataset comprising raster-SVG pairs of slide documents curated from real-world scientific presentations, to facilitate future research in this domain. Our results demonstrate that SliDer achieves a reconstruction LPIPS of 0.069 and is favored by human evaluators in 82.9% of cases compared to the strongest zero-shot VLM baseline.