---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: labels
    dtype:
      class_label:
        names:
          '0': charts
          '1': diagram
          '2': geometry
          '3': medical
          '4': ocr
          '5': random
          '6': table
  splits:
  - name: train
    num_bytes: 160813723527.0
    num_examples: 700768
  - name: test
    num_bytes: 8506367769.25
    num_examples: 36886
  download_size: 169224452489
  dataset_size: 169320091296.25
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---

# 📊 Vision Filtering Dataset

A high-quality, labeled image dataset designed to benchmark computer vision models for filtering noisy image data—especially relevant for pretraining and curating datasets for vision-language models (VLMs).

---

## 📌 Overview

This dataset contains **6 image categories** curated from online and public datasets:

- 📈 `charts`: Graphs, bar charts, line charts, pie charts  
- 🧠 `diagrams`: Schematics, flowcharts, technical illustrations  
- 📐 `geometry`: Geometric shapes, figures, and math visuals  
- 🏥 `medical`: Annotated scans, X-rays, and medical diagrams  
- 🔤 `ocr`: Images containing printed text or handwriting  
- 🌀 `random`: Miscellaneous, non-relevant/noisy images

The dataset is intended for training and evaluating classification models to **automatically filter relevant images** from large-scale scraped datasets.

---
## 🧩 Datasets by Category (with Hugging Face Links)

| Category     | Dataset |
|--------------|---------|
| 📊 **Charts** | [nimapourjafar/mm_chart2text](https://huggingface.co/datasets/nimapourjafar/mm_chart2text) <br> [nimapourjafar/mm_chartqa](https://huggingface.co/datasets/nimapourjafar/mm_chartqa) <br> [nimapourjafar/mm_dvqa](https://huggingface.co/datasets/nimapourjafar/mm_dvqa) <br> [nimapourjafar/mm_figureqa](https://huggingface.co/datasets/nimapourjafar/mm_figureqa) <br> [nimapourjafar/mm_plotqa](https://huggingface.co/datasets/nimapourjafar/mm_plotqa) <br> [nimapourjafar/mm_vistext](https://huggingface.co/datasets/nimapourjafar/mm_vistext) |
| 🔄 **Diagram** | [lmms-lab/ai2d](https://huggingface.co/datasets/lmms-lab/ai2d) <br> [nimapourjafar/mm_tqa](https://huggingface.co/datasets/nimapourjafar/mm_tqa) <br> [shreyanshu09/Block_Diagram](https://huggingface.co/datasets/shreyanshu09/Block_Diagram) <br> [yyyyifan/TQA](https://huggingface.co/datasets/yyyyifan/TQA) |
| 📐 **Geometry** | [5CD-AI/Viet-Geometry-VQA](https://huggingface.co/datasets/5CD-AI/Viet-Geometry-VQA) <br> [AI4Math/MathVerse](https://huggingface.co/datasets/AI4Math/MathVerse) <br> [HuggingFaceM4/datikz](https://huggingface.co/datasets/HuggingFaceM4/datikz) <br> [MathLLMs/MathVision](https://huggingface.co/datasets/MathLLMs/MathVision) <br> [nimapourjafar/mm_geomverse](https://huggingface.co/datasets/nimapourjafar/mm_geomverse) <br> [nimapourjafar/mm_intergps](https://huggingface.co/datasets/nimapourjafar/mm_intergps) <br> [PeijieWang/MV-MATH](https://huggingface.co/datasets/PeijieWang/MV-MATH) <br> [THU-KEG/MM_Math](https://huggingface.co/datasets/THU-KEG/MM_Math) <br> [VIM-Bench/VIM-MathVista](https://huggingface.co/datasets/VIM-Bench/VIM-MathVista) <br> [We-Math/We-Math](https://huggingface.co/datasets/We-Math/We-Math) |
| 🧬 **Medical** | [foreverbeliever/OmniMedVQA](https://huggingface.co/datasets/foreverbeliever/OmniMedVQA) <br> [rbojia/medical-vqa](https://huggingface.co/datasets/rbojia/medical-vqa) |
| 🧾 **OCR** | [5CD-AI/Viet-Geometry-VQA](https://huggingface.co/datasets/5CD-AI/Viet-Geometry-VQA) <br> [mathieu1256/FATURA2-invoices](https://huggingface.co/datasets/mathieu1256/FATURA2-invoices) <br> [nimapourjafar/mm_docvqa](https://huggingface.co/datasets/nimapourjafar/mm_docvqa) <br> [nimapourjafar/mm_iam](https://huggingface.co/datasets/nimapourjafar/mm_iam) <br> [nimapourjafar/mm_ocrvqa](https://huggingface.co/datasets/nimapourjafar/mm_ocrvqa) <br> [nimapourjafar/mm_rendered_text](https://huggingface.co/datasets/nimapourjafar/mm_rendered_text) <br> [nimapourjafar/mm_visualmrc](https://huggingface.co/datasets/nimapourjafar/mm_visualmrc) <br> [nimapourjafar/mm_websight](https://huggingface.co/datasets/nimapourjafar/mm_websight) <br> [vikp/doclaynet_math](https://huggingface.co/datasets/vikp/doclaynet_math) <br> [JayRay5/Image_Infographvqa](https://huggingface.co/datasets/JayRay5/Image_Infographvqa) <br> [nimapourjafar/mm_infographic_vqa](https://huggingface.co/datasets/nimapourjafar/mm_infographic_vqa) <br> [nimapourjafar/mm_finqa](https://huggingface.co/datasets/nimapourjafar/mm_finqa) <br> [nimapourjafar/mm_multihierrt](https://huggingface.co/datasets/nimapourjafar/mm_multihierrt) <br> [nimapourjafar/mm_robust_sqa](https://huggingface.co/datasets/nimapourjafar/mm_robust_sqa) <br> [nimapourjafar/mm_robust_wikisql](https://huggingface.co/datasets/nimapourjafar/mm_robust_wikisql) <br> [nimapourjafar/mm_robust_wtq](https://huggingface.co/datasets/nimapourjafar/mm_robust_wtq) <br> [nimapourjafar/mm_tabmwp](https://huggingface.co/datasets/nimapourjafar/mm_tabmwp) <br> [nimapourjafar/mm_tat_qa](https://huggingface.co/datasets/nimapourjafar/mm_tat_qa) |
| 🌄 **Random** | [COCO Dataset](https://cocodataset.org/) |

---

## 📊 Dataset Distribution

The dataset is well-balanced across six image categories, with slightly more samples in the `ocr` and `random` classes.

<div style="display: flex; justify-content: space-between; align-items: center; flex-wrap: wrap; gap: 20px;">

  <img src="./dataset_distribution_pie_chart.png" alt="Pie Chart - Percentage Distribution" style="width: 35%; height: auto;" />

  <img src="./file_distribution_chart.png" alt="Bar Chart - File Count" style="width: 52%; height: auto;" />

</div>

## 🧾 Dataset Structure

The dataset is organized in a standard image classification folder format:

vision-filtering-dataset/
```
├── train/
│   ├── charts/
│   ├── diagrams/
│   ├── geometry/
│   ├── medical/
│   ├── ocr/
│   └── random/
└── test/
├── charts/
├── diagrams/
├── geometry/
├── medical/
├── ocr/
└── random/
```
Each subfolder contains `.jpg` or `.png` image files.

---

## 🧪 Use Cases

- Vision model training (CNNs, Transformers, ViTs)
- Image filtering for web-scraped datasets
- Preprocessing for multimodal or OCR-based tasks
- Benchmarking classification models on mixed visual domains

---

## 🧠 Loading with 🤗 Datasets

```python
from datasets import load_dataset

dataset = load_dataset("AbdulazizAlshamsi/VLM_Dataset_classification")
train = dataset["train"]
test = dataset["test"]
```

Each sample contains:
	•	image: the image data (PIL object)
	•	label: the class label (charts, diagrams, etc.)

⸻

## 📚 Citation

If you use this dataset, please cite it as follows:

```bibtex
@misc{visionfiltering2025,
  title={Vision Filtering Dataset},
  author={Abdulaziz Alshamsi},
  year={2025},
  howpublished={\url{https://huggingface.co/datasets/AbdulazizAlshamsi/VLM_Dataset_classification}},
  note={Image classification dataset for visual filtering}
}
```
⸻

## 🙋‍♂️ Author

Abdulaziz Alshamsi
AI Researcher — The University of Manchester
📧 abdulaziz.alshamsi@postgrad.manchester.ac.uk
🔗 LinkedIn

⸻

## ❤️ Contributions

Feel free to open issues or submit pull requests to improve the dataset!