Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2401.09220 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866929213070639104 |
|---|---|
| author | Hu, Kai Wang, Jiawei Lin, Weihong Zhong, Zhuoyao Sun, Lei Huo, Qiang |
| author_facet | Hu, Kai Wang, Jiawei Lin, Weihong Zhong, Zhuoyao Sun, Lei Huo, Qiang |
| contents | Existing methods for Visual Information Extraction (VIE) from form-like documents typically fragment the process into separate subtasks, such as key information extraction, key-value pair extraction, and choice group extraction. However, these approaches often overlook the hierarchical structure of form documents, including hierarchical key-value pairs and hierarchical choice groups. To address these limitations, we present a new perspective, reframing VIE as a relation prediction problem and unifying labels of different tasks into a single label space. This unified approach allows for the definition of various relation types and effectively tackles hierarchical relationships in form-like documents. In line with this perspective, we present UniVIE, a unified model that addresses the VIE problem comprehensively. UniVIE functions using a coarse-to-fine strategy. It initially generates tree proposals through a tree proposal network, which are subsequently refined into hierarchical trees by a relation decoder module. To enhance the relation prediction capabilities of UniVIE, we incorporate two novel tree constraints into the relation decoder: a tree attention mask and a tree level embedding. Extensive experimental evaluations on both our in-house dataset HierForms and a publicly available dataset SIBR, substantiate that our method achieves state-of-the-art results, underscoring the effectiveness and potential of our unified approach in advancing the field of VIE. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2401_09220 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | UniVIE: A Unified Label Space Approach to Visual Information Extraction from Form-like Documents Hu, Kai Wang, Jiawei Lin, Weihong Zhong, Zhuoyao Sun, Lei Huo, Qiang Computation and Language Existing methods for Visual Information Extraction (VIE) from form-like documents typically fragment the process into separate subtasks, such as key information extraction, key-value pair extraction, and choice group extraction. However, these approaches often overlook the hierarchical structure of form documents, including hierarchical key-value pairs and hierarchical choice groups. To address these limitations, we present a new perspective, reframing VIE as a relation prediction problem and unifying labels of different tasks into a single label space. This unified approach allows for the definition of various relation types and effectively tackles hierarchical relationships in form-like documents. In line with this perspective, we present UniVIE, a unified model that addresses the VIE problem comprehensively. UniVIE functions using a coarse-to-fine strategy. It initially generates tree proposals through a tree proposal network, which are subsequently refined into hierarchical trees by a relation decoder module. To enhance the relation prediction capabilities of UniVIE, we incorporate two novel tree constraints into the relation decoder: a tree attention mask and a tree level embedding. Extensive experimental evaluations on both our in-house dataset HierForms and a publicly available dataset SIBR, substantiate that our method achieves state-of-the-art results, underscoring the effectiveness and potential of our unified approach in advancing the field of VIE. |
| title | UniVIE: A Unified Label Space Approach to Visual Information Extraction from Form-like Documents |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2401.09220 |