---
dataset_info:
- config_name: absolute
  features:
  - name: sampleId
    dtype: int64
  - name: referenceSentenceId
    dtype: int64
  - name: annotator8
    dtype: int32
  - name: annotator9
    dtype: int32
  - name: annotator10
    dtype: int32
  - name: annotator11
    dtype: int32
  - name: samplePath
    dtype: large_string
  splits:
  - name: test
    num_bytes: 20503
    num_examples: 391
  - name: validation
    num_bytes: 20031
    num_examples: 382
  download_size: 18579
  dataset_size: 40534
- config_name: absolute_pre_study
  features:
  - name: sampleId
    dtype: int64
  - name: referenceSentenceId
    dtype: int64
  - name: annotator41
    dtype: int32
  - name: annotator42
    dtype: int32
  - name: annotator5
    dtype: int32
  - name: annotator6
    dtype: int32
  - name: annotator7
    dtype: int32
  - name: samplePath
    dtype: large_string
  splits:
  - name: train
    num_bytes: 7619
    num_examples: 136
  download_size: 0
  dataset_size: 7619
- config_name: absolute_test
  features:
  - name: sampleId
    dtype: int64
  - name: referenceSentenceId
    dtype: int64
  - name: annotator8
    dtype: int32
  - name: annotator9
    dtype: int32
  - name: annotator10
    dtype: int32
  - name: annotator11
    dtype: int32
  - name: samplePath
    dtype: large_string
  splits:
  - name: train
    num_bytes: 20503
    num_examples: 391
  download_size: 0
  dataset_size: 20503
- config_name: absolute_val
  features:
  - name: sampleId
    dtype: int64
  - name: referenceSentenceId
    dtype: int64
  - name: annotator8
    dtype: int32
  - name: annotator9
    dtype: int32
  - name: annotator10
    dtype: int32
  - name: annotator11
    dtype: int32
  - name: samplePath
    dtype: large_string
  splits:
  - name: train
    num_bytes: 20031
    num_examples: 382
  download_size: 9220
  dataset_size: 20031
- config_name: aspect_ratio
  features:
  - name: samplePath
    dtype: large_string
  - name: sampleId
    dtype: int64
  - name: aspectRatio
    dtype: float64
  splits:
  - name: train
    num_bytes: 40876
    num_examples: 1134
  - name: test
    num_bytes: 14100
    num_examples: 391
  - name: validation
    num_bytes: 13775
    num_examples: 382
  download_size: 46620
  dataset_size: 68751
- config_name: characteristics
  features:
  - name: sampleId
    dtype: int64
  - name: text
    dtype: large_string
  - name: referenceSentenceId
    dtype: int64
  - name: isWrittenInPureCursive
    dtype: bool
  - name: isStrokeThin
    dtype: bool
  - name: containsTypo
    dtype: bool
  - name: containsCorrection
    dtype: bool
  - name: samplePath
    dtype: large_string
  splits:
  - name: train
    num_bytes: 82789
    num_examples: 1134
  - name: test
    num_bytes: 28559
    num_examples: 391
  - name: validation
    num_bytes: 27899
    num_examples: 382
  download_size: 39770
  dataset_size: 139247
- config_name: default
  features:
  - name: userId
    dtype: int64
  - name: sampleId1
    dtype: int64
  - name: sampleId2
    dtype: int64
  - name: referenceSentenceId1
    dtype: int64
  - name: referenceSentenceId2
    dtype: int64
  - name: questionId
    dtype: int64
  - name: score
    dtype: int64
  - name: submissionTimestamp
    dtype: large_string
  - name: samplePath1
    dtype: large_string
  - name: samplePath2
    dtype: large_string
  splits:
  - name: train
    num_bytes: 1395322
    num_examples: 11340
  - name: test
    num_bytes: 481201
    num_examples: 3910
  - name: validation
    num_bytes: 470098
    num_examples: 3820
  download_size: 333435
  dataset_size: 2346621
- config_name: main_study
  features:
  - name: userId
    dtype: int64
  - name: sampleId1
    dtype: int64
  - name: sampleId2
    dtype: int64
  - name: referenceSentenceId1
    dtype: int64
  - name: referenceSentenceId2
    dtype: int64
  - name: questionId
    dtype: int64
  - name: score
    dtype: int64
  - name: submissionTimestamp
    dtype: large_string
  - name: samplePath1
    dtype: large_string
  - name: samplePath2
    dtype: large_string
  splits:
  - name: train
    num_bytes: 1395322
    num_examples: 11340
  - name: test
    num_bytes: 481201
    num_examples: 3910
  - name: validation
    num_bytes: 470098
    num_examples: 3820
  download_size: 333435
  dataset_size: 2346621
- config_name: pre_study
  features:
  - name: userId
    dtype: int64
  - name: sampleId1
    dtype: int64
  - name: sampleId2
    dtype: int64
  - name: referenceSentenceId1
    dtype: int64
  - name: referenceSentenceId2
    dtype: int64
  - name: questionId
    dtype: int64
  - name: score
    dtype: int64
  - name: submissionTimestamp
    dtype: large_string
  - name: samplePath1
    dtype: large_string
  - name: samplePath2
    dtype: large_string
  splits:
  - name: annotator11
    num_bytes: 226487
    num_examples: 1839
  - name: annotator12
    num_bytes: 188442
    num_examples: 1530
  - name: annotator2
    num_bytes: 197791
    num_examples: 1606
  - name: annotator3
    num_bytes: 189783
    num_examples: 1541
  download_size: 102290
  dataset_size: 802503
configs:
- config_name: absolute
  data_files:
  - split: test
    path: absolute/test-*
  - split: validation
    path: absolute/validation-*
- config_name: absolute_pre_study
  data_files:
  - split: train
    path: absolute_pre_study/train-*
- config_name: absolute_test
  data_files:
  - split: train
    path: absolute_test/train-*
- config_name: absolute_val
  data_files:
  - split: train
    path: absolute_val/train-*
- config_name: aspect_ratio
  data_files:
  - split: train
    path: aspect_ratio/train-*
  - split: test
    path: aspect_ratio/test-*
  - split: validation
    path: aspect_ratio/validation-*
- config_name: characteristics
  data_files:
  - split: train
    path: characteristics/train-*
  - split: test
    path: characteristics/test-*
  - split: validation
    path: characteristics/validation-*
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
  - split: validation
    path: data/validation-*
- config_name: main_study
  data_files:
  - split: train
    path: main_study/train-*
  - split: test
    path: main_study/test-*
  - split: validation
    path: main_study/validation-*
- config_name: pre_study
  data_files:
  - split: annotator11
    path: pre_study/annotator11-*
  - split: annotator12
    path: pre_study/annotator12-*
  - split: annotator2
    path: pre_study/annotator2-*
  - split: annotator3
    path: pre_study/annotator3-*
license: mit
language:
- de
tags:
- handwriting
- legibility
- pairwise-comparison
- ranking
- vision
size_categories:
- 10K<n<100K
task_categories:
- text-classification
---

# HaLO: Handwriting Assessment for Legibility and Ordering

HaLO is a dataset of 1,907 handwriting images from 202 German schoolchildren (ages 10--12), annotated with 19,070 pairwise legibility judgments. It supports research on AI-based handwriting legibility assessment through comparative ranking.

## Dataset Description

Each handwriting sample depicts one of ten predefined German sentences (20--27 characters each), written by a child in their natural handwriting style. The samples were scanned and cropped from paper sheets. Due to privacy constraints, only Pixtral 12B image embeddings (1024-dimensional feature vectors) are released, not the raw images.

### Annotation

Legibility annotations were collected using a pairwise comparison paradigm: annotators were shown two handwriting samples side by side and asked to select the more legible one. A pre-study compared this approach with absolute Likert-scale ratings and found that pairwise comparisons produce higher inter- and intra-rater agreement. The main study annotations were collected by eight annotators (three experts, five non-experts).

### Data Splits

The dataset is split into training (1,134 samples, 121 children), validation (382 samples, 40 children), and test (391 samples, 41 children) partitions. All samples from the same child are assigned to the same partition to prevent information leakage.

## Configurations

### `default` / `main_study`

Pairwise legibility annotations from the main study. Each row is a comparison between two samples.

| Field | Description |
|-------|-------------|
| `sampleId1`, `sampleId2` | IDs of the two compared samples |
| `score` | `1` if sample 1 is more legible, `-1` if sample 2 is more legible |
| `samplePath1`, `samplePath2` | Paths to the corresponding feature `.npy` files |
| `userId` | Annotator ID |
| `referenceSentenceId1`, `referenceSentenceId2` | Reference sentence IDs (1--10) |
| `questionId` | Annotation question ID |
| `submissionTimestamp` | Timestamp of the annotation |

**Splits:** `train` (11,340), `test` (3,910), `validation` (3,820)

```python
from datasets import load_dataset
dataset = load_dataset("MarcoLents/HaLO")
```

### `pre_study`

Pairwise legibility annotations from the annotation pre-study, with the same schema as the main study. Splits correspond to individual annotators, where Annotator 1 performed the procedure twice (splits `annotator11` and `annotator12`).

**Splits:** `annotator11` (1,839), `annotator12` (1,530), `annotator2` (1,606), `annotator3` (1,541)

```python
pre_study = load_dataset("MarcoLents/HaLO", "pre_study")
```

### `characteristics`

Binary sample characteristics annotated per sample.

| Field | Description |
|-------|-------------|
| `sampleId` | Sample ID |
| `text` | Transcribed text content |
| `referenceSentenceId` | Reference sentence ID |
| `isWrittenInPureCursive` | Whether the sample is written entirely in cursive |
| `isStrokeThin` | Whether the stroke width is predominantly thin |
| `containsTypo` | Whether at least one typographical error is present |
| `containsCorrection` | Whether at least one correction is present |
| `samplePath` | Path to the feature `.npy` file |

**Splits:** `train` (1,134), `test` (391), `validation` (382)

```python
characteristics = load_dataset("MarcoLents/HaLO", "characteristics")
```

### `absolute_pre_study`

Absolute legibility ratings on a 5-point Likert scale (0 = very legible, 4 = not legible at all) from the annotation pre-study. Four annotators independently rated a random subset of 136 samples. Annotator 4 rated twice (`annotator41`, `annotator42`).

| Field | Description |
|-------|-------------|
| `sampleId` | Sample ID |
| `referenceSentenceId` | Reference sentence ID |
| `annotator41` -- `annotator7` | Integer Likert ratings (0--4) per annotator |
| `samplePath` | Path to the feature `.npy` file |

**Splits:** `train` (136)

```python
absolute_pre_study = load_dataset("MarcoLents/HaLO", "absolute_pre_study")
```

### `absolute_test`

Absolute legibility ratings on a 5-point Likert scale (0 = very legible, 4 = not legible at all) from the test set. Four annotators independently rated samples from the test partition. Not all annotators rated every sample; missing values indicate unrated samples. Where a rater provided multiple ratings for the same sample, the value is the rounded mean.

> **Note:** The `absolute_pre_study` and `absolute_test` configs use different annotator numbering schemes because they originate from separate annotation campaigns described in different sections of the paper. Some of the underlying raters overlap between the two configs, but the pseudonyms are independent.

| Field | Description |
|-------|-------------|
| `sampleId` | Sample ID |
| `referenceSentenceId` | Reference sentence ID |
| `annotator8` -- `annotator11` | Integer Likert ratings (0--4) per annotator, NaN if not rated |
| `samplePath` | Path to the feature `.npy` file |

**Splits:** `train` (391)

```python
absolute_test = load_dataset("MarcoLents/HaLO", "absolute_test")
```

### `aspect_ratio`

Image aspect ratio for each sample.

| Field | Description |
|-------|-------------|
| `sampleId` | Sample ID |
| `aspectRatio` | Width-to-height ratio of the original image |
| `samplePath` | Path to the feature `.npy` file |

**Splits:** `train` (1,134), `test` (391), `validation` (382)

```python
aspect_ratio = load_dataset("MarcoLents/HaLO", "aspect_ratio")
```

## Feature Files

Pixtral 12B image embeddings are stored as `.npy` files under `features/{referenceSentenceId}/{sampleId}.npy`. Each file contains a 1024-dimensional float32 vector obtained by mean-pooling the token-level outputs of the frozen Pixtral-ViT encoder.

To download the feature files:

```python
from huggingface_hub import snapshot_download
snapshot_download(
    repo_id="MarcoLents/HaLO",
    repo_type="dataset",
    allow_patterns="features/*/*.npy",
    local_dir="data"
)
```

## Citation

If you use this dataset, please cite:

```bibtex
@article{bauer2025halo,
  title={Ranking handwriting images like a human: AI-based legibility assessment by comparative ranking on the HaLO dataset},
  author={Bauer, Meike and Lents, Marco and Schmidt, Erik and Hamann, Tim and Pieger, Lukas and Salata, Susanne and Di Salvo, Francesco and Hoffmann, Tal and Barth, Jens and Ledig, Christian},
  year={2025}
}
```

## License

This dataset is released under the [MIT License](LICENSE).

## Acknowledgements

This study was supported by the Hightech Agenda Bayern (HTA) of the Free State of Bavaria, Germany. Additional support was provided by the Bavarian State Ministry of Economic Affairs, Regional Development and Energy (StMWi) as part of the Bavarian Collaborative Research Programme (BayVFP), funding line Digitalization, project KIBEL, grant number DIK0813.