README.md

---
license: cc-by-nd-4.0
language:
- fa
task_categories:
- image-to-text
- text-to-image
tags:
- ocr
- synthetic
- persian
- farsi
- naskh
- document-ai
pretty_name: Persian Pixel
size_categories:
- 100K<n<1M
configs:
- config_name: sentence
  data_files:
  - split: train
    path: data/sentence-*.parquet
- config_name: paragraph
  data_files:
  - split: train
    path: data/paragraph-*.parquet
- config_name: page
  data_files:
  - split: train
    path: data/page-*.parquet
- config_name: full
  data_files:
  - split: train
    path:
    - data/sentence-*.parquet
    - data/paragraph-*.parquet
    - data/page-*.parquet
---

# Persian Pixel

**Persian Pixel** is a synthetic optical character recognition (OCR) dataset for **Persian / Farsi (`fa`)**, in which Unicode text is rendered to images and paired with its exact transcription. It is built for OCR recognition, image-to-text modeling, fine-tuning, and evaluation workflows that need clean, controllable image/label pairs at scale.

Because the text is rendered programmatically, every image ships with a perfectly aligned ground-truth label — making the dataset well suited for pretraining, curriculum learning, and isolating model behavior before introducing noisy real-world scans.


**Authors:** [Haq Nawaz Malik](https://huggingface.co/Omarrran) · [PouriaMahdi84](https://huggingface.co/PouriaMahdi84)

### Sentences

| | | | |
|---|---|---|---|
| <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/DaUvNkGNwIg8WWN_p_8tI.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/tnB2Mm4szGrH9aXQR59rh.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/5hp_TrKyM2SRegVP_YNOq.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/adsv_GMKAN2ZmFG9BkjfM.png" width="220"> |
| <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/NoCHc0RQqIO0GBtY3ihf6.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/TYo56JUOnxJWZCVpoiS8g.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/pog584pXTsNGxjakBrptL.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/XejD-RDT9L7PnFh71wBe3.png" width="220"> |
| <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/osTMjKu8qLoh2HMRfLU04.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/xGZGOmqQ05H2OZXcSqL4d.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/sHSSLObetkLYqIhmRzpLu.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/VPp8wFA4whw_d1GZ-3CM9.png" width="220"> |



### Paragraphs

| | | | |
|---|---|---|---|
| <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/QjoEM3E1rax5WnMr898Qb.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/PT8lfxu2peI3D4VQp_ZNW.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/7E8X_GP5NXIP3Rgga-d-u.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/_8KY-hjAi2xk6Uq8x4odm.png" width="220"> |
| <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/X9n14KHjWiIw4zWLJZrik.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/RMGuPbWlyO0zcMP1-LBz0.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/YuOPUx8uIazmwLcTx8ejU.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/1peSzR7out6Y_vdb_7F0r.png" width="220"> |
| <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/5CB9Nw0UFXkJA8MwuUfsw.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/at_0aB7_HfOFJeSoXPaxC.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/MUwCrT-Ao_z936skgpU3-.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/nFUCpAmwPTYAOtH4WWC6D.png" width="220"> |




### Pages

| | | | |
|---|---|---|---|
| <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/KYeuz-Xr2oXVM5DXyPWRq.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/3J6adzWLvZktaNRO7oUp2.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/zf_2Shk-gRwkNk_cwQ_wS.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/0JVX6ZauIW8Zd4wLsrPgO.png" width="220"> |
| <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/lhOSV90M4KldIYDxUREm2.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/gtvYQkgeg-bJ0_Gq3smnX.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/ZEysUBC0oTaXXVJE_pprk.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/C7Tm-hE_HuMbnnBsEzpjs.png" width="220"> |
| <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/9h6agmJn5wZoIz8i_i0od.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/PeY29MfMKeafQ6Rabijuz.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/QWqGVo9Z-yL7na7_1eW6g.png" width="220"> | <img src="https://cdn-uploads.huggingface.co/production/uploads/66afb3f1eaf3e876595627bf/FAzH6OeSabiDKlEYhI9sd.png" width="220"> |


## Dataset summary

| Property | Value |
|---|---|
| Language | Persian / Farsi (`fa`) |
| Script | Perso-Arabic |
| Modality | Image → Text (OCR) |
| Source | Synthetic (font-rendered) |
| Total rows (`full`) | 343,246 |
| License | CC BY-ND 4.0 |

## Configurations and sizes

The dataset is organized by render granularity. Each granularity is exposed as its own config, plus a `full` config that virtually concatenates all three.

| Config | Description | Rows | Typical width |
|---|---|---:|---|
| `sentence` | Short line / sentence-level crops |251,000 | 640 px |
| `paragraph` | Multi-line paragraph blocks | 110,138 | 512 px |
| `page` | Page-level renders | 31,108 | 768 px |
| `full` | Concatenation of all three configs above | 343,246 | — |

Verified from Parquet shard metadata on **2026-06-25 05:34:30 UTC** (320 Parquet files, ~31.8 GB).

### Page-config statistics

| Metric | Min | Max | Mean |
|---|---:|---:|---:|
| Height (px) | 38 | 1100 | 597.92 |
| Text length (chars) | 13 | 1896 | 824.22 |
| Line count | 2 | 75 | 7.88 |

## Data fields

Every row shares the same schema across all configs:

| Field | Type | Description |
|---|---|---|
| `bytes` | `bytes` | Encoded image bytes |
| `path` | `string` | Image path/name stored with the image bytes |
| `text` | `string` | Exact Persian Unicode OCR target / transcription |
| `sample_type` | `string` | One of `sentence`, `paragraph`, or `page` |
| `source_run_id` | `string` | Internal generation / publish run identifier |
| `image_path` | `string` | Original relative image path |
| `width` | `int` | Rendered image width in pixels |
| `height` | `int` | Rendered image height in pixels |
| `text_chars` | `int` | Number of Unicode characters in `text` |
| `line_count` | `int` | Number of text lines in the label |

## Usage

Install the dependency:

```bash
pip install datasets pillow
```

Load a single config:

```python
from datasets import load_dataset

train = load_dataset("Omarrran/Persian_Pixel", "sentence", split="train")  # or "paragraph" / "page"

print(train.column_names)
print(train[0]["text"])
```

Load the combined config:

```python
from datasets import load_dataset

full = load_dataset("Omarrran/Persian_Pixel", "full", split="train")
print(len(full))  # 343246
```

Stream a config (recommended for the heavier `paragraph` / `page` renders):

```python
from datasets import load_dataset

ds = load_dataset(
    "Omarrran/Persian_Pixel",
    name="page",
    split="train",
    streaming=True,
)

for sample in ds:
    txt = sample["text"]
    # feed into OCR preprocessing / tokenizer pipeline
```

If your loader sees raw `bytes` and `path` columns rather than a decoded `image` object, reconstruct a PIL image:

```python
from io import BytesIO
from PIL import Image

row = train[0]
img = Image.open(BytesIO(row["bytes"]))
print(img.size, row["text"])
```

### Loading each config

```python
from datasets import load_dataset

sentence  = load_dataset("Omarrran/Persian_Pixel", "sentence",  split="train")
paragraph = load_dataset("Omarrran/Persian_Pixel", "paragraph", split="train")
page      = load_dataset("Omarrran/Persian_Pixel", "page",      split="train")
full      = load_dataset("Omarrran/Persian_Pixel", "full",      split="train")
```

## Recommended training split strategy

The repo publishes data as `train` shards. For model development, create your own deterministic split:

```python
ds = load_dataset("Omarrran/Persian_Pixel", "sentence", split="train")
split = ds.train_test_split(test_size=0.02, seed=42)
train_ds, val_ds = split["train"], split["test"]
```

For page-level OCR, keep validation smaller and representative because page images are much heavier than sentence crops.

## Recommended use cases

- Fine-tuning image-to-text and OCR models such as TrOCR, Donut, BLIP-2, and PaliGemma-style decoders.
- Pretraining or curriculum stages before introducing noisy scanned or photographed OCR corpora.
- Evaluating normalization, decoding, and post-correction pipelines on Perso-Arabic script.
- Benchmarking Persian text recognition under controlled synthetic conditions and variable line counts.

## Limitations and considerations

- This is **synthetic** OCR data. Performance on font-rendered text does not guarantee performance on real scans, photographs, or handwriting — validate on in-domain data before any production deployment.
- Synthetic rendering can introduce artifacts, unusual line wrapping, punctuation variation, or source-text noise.
- The `paragraph` and `page` configs contain long, multi-line labels and large images; always consume them through the `datasets` library rather than raw parsing.

## License

Released under **Creative Commons Attribution-NoDerivatives 4.0 International (CC BY-ND 4.0)**. You may copy and redistribute the dataset for any purpose with attribution, but you may not distribute modified versions.

## Citation

```bibtex
@dataset{persian_pixel_2026,
  title     = {Persian Pixel: A Synthetic OCR Dataset for Persian/Farsi},
  author    = { Haq Nawaz Malik, Pouria Mahdi},
  year      = {2026},
  publisher = {Hugging Face},
  url       = {https://huggingface.co/datasets/Omarrran/Persian_Pixel}
}
```
## Repository

- Dataset repo: https://huggingface.co/datasets/Omarrran/Persian_Pixel
- Format: Parquet shards
- Primary task: Persian OCR / image-to-text