|
--- |
|
license: mit |
|
datasets: |
|
- fimu-docproc-research/CIVQA-TesseractOCR |
|
language: |
|
- cs |
|
tags: |
|
- document question answering |
|
--- |
|
|
|
# LayoutXLM Model Fine-tuned with CIVQA (Tesseract) dataset |
|
|
|
This is a fine-tuned version of the [LayoutXLM model](https://huggingface.co/microsoft/layoutxlm-base), which was trained on Czech Invoice Visual Question Answering (CIVQA) datasets containing invoices in the Czech language. |
|
|
|
This model enables Document Visual Question Answering on Czech invoices. |
|
|
|
All invoices used in this dataset were obtained from public sources. Over these invoices, we were focusing on 15 different entities, which are crucial for processing the invoices. |
|
- Invoice number |
|
- Variable symbol |
|
- Specific symbol |
|
- Constant symbol |
|
- Bank code |
|
- Account number |
|
- ICO |
|
- Total amount |
|
- Invoice date |
|
- Due date |
|
- Name of supplier |
|
- IBAN |
|
- DIC |
|
- QR code |
|
- Supplier's address |
|
|
|
You can find more information about this model in this [paper](https://nlp.fi.muni.cz/raslan/raslan23.pdf#page=31). |
