AIbase
EN
Home
CIVQA Impira QA EasyOCR
C

CIVQA Impira QA EasyOCR

Developed by Sharka
This model is a fine-tuned version of the Impira model, specifically designed for visual question answering tasks on Czech invoices.
Text-to-Image
Transformers
Open Source License:MIT#Czech invoice parsing#Multi-field automatic extraction#Document visual question answering
Downloads 74
Release Time : 12/18/2023

Model Overview

Based on Impira's LayoutLM document QA model, this version is fine-tuned using the Czech Invoice Visual Question Answering (CIVQA) dataset, supporting recognition and Q&A for 15 key fields in Czech invoices.

Model Features

Czech invoice processing
Visual question answering capabilities optimized specifically for Czech invoices
15 key field recognition
Accurately identifies 15 key fields including invoice number, bank code, total amount, etc.
Trained on public data
Utilizes Czech invoice data from public sources for training

Model Capabilities

Invoice text recognition
Visual question answering
Multi-field extraction
Czech document processing

Use Cases

Financial automation
Invoice information extraction
Automatically extracts key information from Czech invoices
Accurately identifies 15 key fields
Accounting system integration
Automatically inputs extracted invoice data into accounting systems
Reduces manual data entry workload
Document processing
Invoice Q&A system
Queries invoice information through natural language questions
Quickly retrieves specific invoice field information
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