achimrabus/crnn-ctc-church-slavonic

Church Slavonic HTR Model (Puigcerver CRNN)

A Handwritten Text Recognition (HTR) model for Church Slavonic manuscripts, based on the CNN + BiLSTM + CTC architecture introduced in Puigcerver (2017) and used as the backbone of PyLaia and Transkribus.

This is a clean-room PyTorch reimplementation of that published architecture (PyLaia-inspired). It does not use the PyLaia Python package and is not loadable by it — training and inference run via plain PyTorch (see Usage below).

Model Details

• Architecture: CNN encoder [12, 24, 48, 48 filters] + 3-layer Bidirectional LSTM (256 units) + CTC decoder (Puigcerver 2017)
• Input: Grayscale line images, normalized to 128 px height with aspect ratio preserved
• Output: UTF-8 text (Church Slavonic characters including titlos, abbreviation marks, and diacritics)
• Vocabulary: 152 symbols (symbols.txt)
• Framework: Pure PyTorch — clean-room reimplementation of the Puigcerver (2017) architecture (PyLaia-inspired); the PyLaia package is not required

Performance

Metric	Value
Validation CER	2.89%
Training epochs	59
Training lines	309,959
Training pages	2,643
Validation lines	20,679
Validation pages	205

Training Data

Trained on Church Slavonic handwriting images transcribed and exported from Transkribus (see the corresponding Transkribus model page). The dataset covers Old Cyrillic script styles (uncial and semi-uncial), primarily East Slavic with South Slavic material included.

Source manuscripts:

• Codex Suprasliensis (10th–11th c., South Slavic recension)
• Catecheses of Cyril of Jerusalem (transmitted version: 11th c., East Slavic recension)
• Methodius of Olympus: Symposion (transmitted version: 17th c., East Slavic recension)
• Velikie Minei Četʹi (16th c., East Slavic recension): large parts of the March and May volumes, Apostolos from the June volume

Our CRNN-CTC model was trained on the full collection: 309,959 training lines (2,643 pages) and 20,679 validation lines (205 pages), exported from Transkribus.

The Transkribus model was trained by Elena Renje as part of the QuantiSlav project and curated by Achim Rabus (Slavic Department, University of Freiburg).

Usage

Requirements

pip install torch torchvision pillow

Inference

Download best_model.pt, symbols.txt, and model_config.json from this repository, then use the inference script from polyscriptor:

from inference_pylaia_native import PyLaiaInference
from PIL import Image

# Load model
model = PyLaiaInference(
    checkpoint_path="best_model.pt",
    syms_path="symbols.txt"
)

# Transcribe a line image
image = Image.open("line_image.jpg")
text = model.transcribe(image)
print(text)

Note: Input should be a single text line image, not a full page. Preprocessing (grayscale conversion, height normalization, aspect ratio preservation) is handled automatically by inference_pylaia_native.py.

For full-page inference with automatic line segmentation, use batch_processing.py:

python batch_processing.py \
    --engine crnn-ctc \
    --model-path best_model.pt \
    --input-folder images/ \
    --output-folder output/

GUI Usage

polyscriptor also ships graphical interfaces that handle full-page processing without requiring pre-segmented line images:

Interactive single-page GUI — loads raw page images, performs automatic line segmentation, and can export results as PAGE XML:

python transcription_gui_plugin.py

Batch processing GUI — processes entire folders; auto-detects existing PAGE XML files (e.g. from Transkribus) and uses them for segmentation when available:

python polyscriptor_batch_gui.py

Intended Use

• Transcription of Church Slavonic historical manuscripts
• Research in Slavic medieval studies and digital humanities

Limitations

• Full-page segmentation quality depends on the segmentation method used upstream

Citation

If you use this model in your research, please cite the architecture paper and this model:

@article{puigcerver2017multidimensional,
  title     = {Are Multidimensional Recurrent Layers Really Necessary for Handwritten Text Recognition?},
  author    = {Puigcerver, Joan},
  journal   = {Proceedings of the 14th IAPR International Conference on Document Analysis and Recognition (ICDAR)},
  year      = {2017},
  url       = {https://www.jpuigcerver.net/pubs/jpuigcerver_icdar2017.pdf}
}

@misc{rabus2026polyscriptor,
  title  = {Polyscriptor: Multi-Engine HTR Training \& Comparison Tool},
  author = {Rabus, Achim},
  year   = {2026},
  url    = {https://github.com/achimrabus/polyscriptor}
}