AIbase
EN
Home
Bros Large Uncased
B

Bros Large Uncased

Developed by naver-clova-ocr
BROS is a pre-trained language model focusing on text and layout, designed to better extract key information from documents.
Large Language Model
Transformers
#Document Key Information Extraction#OCR Enhancement Processing#Spatial Relationship Modeling
Downloads 55
Release Time : 3/2/2022

Model Overview

BROS (BERT-based Representation of Spatial Relations) is a pre-trained model that combines text and spatial layout information, specifically designed for document key information extraction tasks, such as extracting ordered item lists from receipts.

Model Features

Text and Layout Integration
Simultaneously considers text content and spatial layout information to enhance document understanding.
Pre-trained Model
Pre-trained on large-scale document data, fine-tunable for specific tasks.
Spatial Relationship Modeling
Specifically designed to capture spatial relationships between document elements.

Model Capabilities

Document Key Information Extraction
Joint Understanding of Text and Layout
Receipt Information Parsing
Table Data Extraction

Use Cases

Document Processing
Receipt Information Extraction
Extracts itemized lists, prices, and other information from scanned receipts.
Accurately identifies structured information in receipts.
Table Data Extraction
Extracts table content from documents while preserving structure.
Maintains original row and column relationships of tables.
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
English简体中文繁體中文にほんご
© 2025AIbase