TDTU VQA Dataset — Vietnamese Medicinal Herbs 🌿

Dataset Description

TDTU VQA Dataset Herb is a Vietnamese Visual Question Answering (VQA) dataset focused on medicinal plants and herbs. It was developed for scientific research at Ton Duc Thang University (TDTU), with the goal of advancing AI models capable of recognizing and answering questions about Vietnamese medicinal herbs.

• Homepage: Hugging Face Dataset
• Repository: azan100an/tdtu_vqa_dataset_herb
• Point of Contact: Ton Duc Thang University (TDTU)

Dataset Summary

Dataset Structure

Splits

Data Fields

Each sample contains 3 fields:

Sample

{
    "file_name": "data/P000000001.jpg",
    "question": "Tên dược liệu trong ảnh là gì?",              # "What is the name of the medicinal herb in the image?"
    "answer":   "Tên dược liệu trong ảnh là Cao cẳng bắc bộ." # "The herb in the image is Cao cẳng bắc bộ."
}

Directory Structure

tdtu_vqa_dataset_herb/
├── data/
│   ├── P000000001.jpg
│   ├── P000000002.jpg
│   └── ... (3,790 images)
├── train_metadata.jsonl   (23,271 lines)
├── val_metadata.jsonl     (2,954 lines)
└── test_metadata.jsonl    (2,926 lines)

Dataset Creation

Source Data

The dataset was compiled from Vietnamese medicinal herb references and botanical resources. Images depict various parts of medicinal plants — leaves, flowers, fruits, roots, bark, and dried preparations — and Q&A pairs are constructed to cover multiple aspects of each herb.

Question Types

The dataset covers a wide range of question types:

Data Processing

• One item with a missing image was excluded (P000001388)
• Data was randomly shuffled with seed=42 for reproducibility
• All text is encoded in UTF-8 with full Vietnamese character support

Usage

Loading the Dataset

from datasets import load_dataset

dataset = load_dataset("azan100an/tdtu_vqa_dataset_herb")

train_data = dataset["train"]
val_data   = dataset["validation"]
test_data  = dataset["test"]

print(train_data[0])
# {
#   "file_name": "data/P000001116.jpg",
#   "question":  "Màu sắc chủ đạo của lá cây trong ảnh là gì?",
#   "answer":    "Màu sắc chủ đạo của lá cây là màu xanh lục với các đường gân lá nổi rõ."
# }

Loading with Images

from datasets import load_dataset
from PIL import Image

dataset = load_dataset("azan100an/tdtu_vqa_dataset_herb")

sample = dataset["train"][0]
image  = Image.open(sample["file_name"])
print("Q:", sample["question"])
print("A:", sample["answer"])

Intended Uses

Appropriate Uses ✅

• Training and evaluating Visual Question Answering (VQA) models for Vietnamese
• Research on medicinal herb recognition using computer vision
• Developing Vietnamese medical/pharmaceutical chatbots
• Academic research on Multimodal AI in the healthcare domain

Out-of-Scope Uses ❌

• Clinical medical diagnosis
• Replacing advice from qualified healthcare professionals
• Commercial applications without explicit permission

Limitations

• The dataset is Vietnamese only and does not support other languages
• Answers are derived from reference materials and may contain domain inaccuracies
• Medical information in this dataset should not be used as a basis for treatment decisions
• The dataset has not been independently verified by licensed pharmacists or botanists

Citation

If you use this dataset in your research, please cite:

@dataset{tdtu_vqa_herb_2025,
  author    = {azan100an and Ton Duc Thang University},
  title     = {TDTU VQA Dataset: Vietnamese Medicinal Herbs},
  year      = {2025},
  publisher = {Hugging Face},
  url       = {https://huggingface.co/datasets/azan100an/tdtu_vqa_dataset_herb}
}

License

This dataset is released under the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) license.

You are free to use, share, and adapt the dataset for non-commercial purposes, provided that appropriate credit is given.

Built with ❤️ at Ton Duc Thang University (TDTU), Ho Chi Minh City, Vietnam