DocSynthv2: A Practical Autoregressive Modeling for Document Generation

TL;DR

DocSynthv2 introduces an autoregressive model that jointly generates document layout and content, advancing the field of comprehensive document creation without visual cues, and demonstrating improved quality and relevance in generated documents.

Contribution

The paper presents a novel autoregressive model that integrates layout and textual cues for comprehensive document generation, surpassing existing layout-only methods.

Findings

Enhanced document generation quality and relevance

Effective handling of layout and textual content integration

Demonstrated superiority on a curated benchmark

Abstract

While the generation of document layouts has been extensively explored, comprehensive document generation encompassing both layout and content presents a more complex challenge. This paper delves into this advanced domain, proposing a novel approach called DocSynthv2 through the development of a simple yet effective autoregressive structured model. Our model, distinct in its integration of both layout and textual cues, marks a step beyond existing layout-generation approaches. By focusing on the relationship between the structural elements and the textual content within documents, we aim to generate cohesive and contextually relevant documents without any reliance on visual components. Through experimental studies on our curated benchmark for the new task, we demonstrate the ability of our model combining layout and textual information in enhancing the generation quality and relevance of documents, opening new pathways for research in document creation and automated design. Our findings emphasize the effectiveness of autoregressive models in handling complex document generation tasks.