---
license: cc-by-nc-4.0
task_categories:
- image-feature-extraction
- object-detection
language:
- vi
- en
tags:
- document-ai
- table-structure-recognition
- table-recognition
- ocr
- financial-reports
- vietnamese
pretty_name: VietFinTab
size_categories:
- 10K<n<100K
---

# VietFinTab

## Dataset Overview

**VietFinTab** is a large-scale Vietnamese financial table dataset designed for **Table Structure Recognition (TSR)**, **Document AI**, and **Optical Character Recognition (OCR)** research.

The dataset is collected from publicly available financial reports of **13 Vietnamese companies** spanning the period from **2023 to 2025**.

The companies included in the dataset are:

* EVN Finance Joint Stock Company (**EVF**)
* FPT Corporation (**FPT**)
* Hoang Anh Gia Lai Joint Stock Company (**HAG**)
* Hoa Phat Group Joint Stock Company (**HPG**)
* Military Commercial Joint Stock Bank (**MB**)
* PetroVietnam Drilling & Well Services Corporation (**PVD**)
* PetroVietnam Technical Services Corporation (**PVS**)
* SSI Securities Corporation (**SSI**)
* TPBank  (**TPB**)
* Vingroup Joint Stock Company (**VIN**)
* Vietjet Aviation Joint Stock Company (**VJC**)
* Vinamilk (**VNM**)
* VietinBank (**VTB**)

---

## Dataset Statistics

* **10,383** table images (`.jpg`)
* **10,383** table structure annotations (`.xml`)
* **1,000** table content annotations (`.json`)

Each table image has a corresponding XML file describing its table structure. Additionally, 1,000 tables include detailed cell text annotations with bounding boxes.

---

## Directory Structure

```text
VietFinTab/
│
├── images/
│   ├── EVF/
│   ├── FPT/
│   ├── ...
│
├── structures/
│   ├── EVF/
│   ├── FPT/
│   ├── ...
│
├── contents/
│   ├── EVF/
│   ├── FPT/
│   ├── ...
│
└── metadata.jsonl
```

---

## File Formats

### Table Images

* Format: JPG
* One image corresponds to one table extracted from a financial report.

### Table Structure Annotation

Each XML annotation follows the **PubTables-1M annotation format**, containing:

* Table bounding box
* Row bounding boxes
* Column bounding boxes
* Spanning cell annotations

These annotations are suitable for training and evaluating table structure recognition models.

### Table Content Annotation

For 1,000 selected tables, JSON annotations provide:

* Text
* Text bounding boxes

These annotations enable OCR-related tasks, text recognition, and end-to-end table understanding.

---

## Metadata

The dataset includes a `metadata.jsonl` file.

Example:

```json
{
  "id": "FPT_Q1_2023_0001",
  "company": "FPT",
  "quarter": "Q1",
  "year": 2023,
  "file_name": "images/FPT/FPT_Q1_2023_0001.jpg",
  "structure_file": "structures/FPT/FPT_Q1_2023_0001.xml",
  "content_file": "contents/FPT/FPT_Q1_2023_0001.json",
  "has_content_label": true
}
```

---

## Supported Tasks

VietFinTab can be used for:

* Table Structure Recognition (TSR)
* Table Detection
* Table Parsing
* OCR
* End-to-End Table Understanding
* Financial Document Analysis

---

## Source

The dataset is constructed from publicly available financial reports released by Vietnamese companies.

The original reports are used solely for academic research and dataset construction.

---

## Citation

If you use VietFinTab in your research, please cite the accompanying publication.

```bibtex
@dataset{vietfintab,
  title={VietFinTab: A Vietnamese Financial Table Dataset},
  author={Pham Hong Quan},
  year={2026},
  publisher={Hugging Face},
  url={https://huggingface.co/datasets/WuanHP/VietFinTab}
}
```

---

## License

This dataset is released under the **CC BY-NC 4.0** license.

It may be used for **research and non-commercial purposes**. Users are responsible for complying with the licensing terms of the original financial reports from which the tables were extracted.