README.md

---
license: mit
language:
- la
- fr
task_categories:
- image-to-text
pretty_name: Himanis
dataset_info:
  features:
  - name: image
    dtype: image
  - name: text
    dtype: string
  splits:
  - name: train
    num_examples: 18504
  - name: validation
    num_examples: 2367
  - name: test
    num_examples: 2240
  dataset_size: 23111
tags:
  - atr
  - ocr
  - htr
  - historical
  - handwritten
---

# Himanis - line level

## Table of Contents
- [Himanis - line level](#himanis-line-level)
  - [Table of Contents](#table-of-contents)
  - [Dataset Description](#dataset-description)
    - [Languages](#languages)
  - [Dataset Structure](#dataset-structure)
    - [Data Instances](#data-instances)
    - [Data Fields](#data-fields)

## Dataset Description

- **Homepage:** [Himanis](http://himanis.huma-num.fr/app//)
- **Paper:** [Paper](https://doi.org/10.4000/medievales.8198)
- **Point of Contact:** [TEKLIA](https://teklia.com)

## Dataset Summary 

Himanis (HIstorical MANuscript Indexing for user controlled Search) is a corpus of medieval documents.

### Languages

All the documents in the dataset are written in Latin and in French.

## Dataset Structure

### Data Instances

```
{
  'image': <PIL.JpegImagePlugin.JpegImageFile image mode=RGB size=4300x128 at 0x1A800E8E190,
  'text': 'Philippus, Dei gratia Francorum et Navarre rex. Notum facimus universis, tam presentibus quam futuris, quod, cum supplicato nobis nuper,'
}
```

### Data Fields


- `image`: A PIL.Image.Image object containing the image. Note that when accessing the image column (using dataset[0]["image"]), the image file is automatically decoded. Note that all images are resized to a fixed height of 128 pixels. Decoding of a large number of image files might take a significant amount of time. Thus it is important to first query the sample index before the "image" column, i.e. dataset[0]["image"] should always be preferred over dataset["image"][0].
- `text`: the label transcription of the image.