DocFusion: A Unified Framework for Document Parsing Tasks

ACL ARR 2025 February Submission4104 Authors

Abstract: Document parsing involves layout element detection and recognition, essential for analyzing complex structures and extracting key information. However, existing methods often employ multiple models for these tasks, leading to increased system complexity and maintenance overhead. While some models attempt to unify detection and recognition, they often fail to address the intrinsic differences in data representations, thereby limiting performance in document processing. Our research reveals that recognition relies on discrete tokens, whereas detection relies on continuous coordinates, leading to challenges in gradient updates and optimization. To bridge this gap, we propose the Gaussian-Kernel Cross-Entropy Loss (GK-CEL), enabling generative frameworks to handle both tasks simultaneously. Building upon GK-CEL, we propose DocFusion, a unified document parsing model with only 0.28B parameters. Additionally, we construct the DocLatex-1.6M dataset to provide high-quality training support. Experimental results demonstrate that DocFusion leverage GK-CEL effectively exploits the benefits of multi-task learning and achieves state-of-the-art performance across four key tasks.

Paper Type: Long

Research Area: Multimodality and Language Grounding to Vision, Robotics and Beyond

Research Area Keywords: Multimodality and Language Grounding to Vision, Robotics and Beyond,Information Extraction

Contribution Types: NLP engineering experiment, Approaches to low-resource settings, Data resources

Languages Studied: English

Submission Number: 4104