FS-DAG: Few Shot Domain Adapting Graph Networks for Visually Rich Document Understanding

TL;DR

FS-DAG is a scalable, efficient few-shot model for visually rich document understanding that adapts to diverse document types with minimal data, handling OCR errors and domain shifts effectively.

Contribution

Introduces FS-DAG, a modular, low-parameter model architecture for VRDU that improves performance and robustness in few-shot settings compared to existing methods.

Findings

Significant improvements in convergence speed and accuracy.

Robustness to OCR errors and domain shifts.

Achieves high performance with less than 90M parameters.

Abstract

In this work, we propose Few Shot Domain Adapting Graph (FS-DAG), a scalable and efficient model architecture for visually rich document understanding (VRDU) in few-shot settings. FS-DAG leverages domain-specific and language/vision specific backbones within a modular framework to adapt to diverse document types with minimal data. The model is robust to practical challenges such as handling OCR errors, misspellings, and domain shifts, which are critical in real-world deployments. FS-DAG is highly performant with less than 90M parameters, making it well-suited for complex real-world applications for Information Extraction (IE) tasks where computational resources are limited. We demonstrate FS-DAG's capability through extensive experiments for information extraction task, showing significant improvements in convergence speed and performance compared to state-of-the-art methods. Additionally, this work highlights the ongoing progress in developing smaller, more efficient models that do not compromise on performance. Code : https://github.com/oracle-samples/fs-dag