virtualenv -p python3 venv
source venv/bin/activate
pip install -r requirements.txt

source venv/bin/activate
python run_iam.py --model=model/best_lm=0.5_beams=5/serve --data=data/test.lst

In src we share our implementations of the beam-search and the CTC-prefix-scores. Note, that the full algorithm is build in the Tensorflow graph which is why a simple model.predict is sufficient to obtain the decoded sequence.

Please cite

Wick, C., Zöllner, J., and Grüning, T., "Rescoring Sequence-to-Sequence Models for Text Line Recognition with CTC-Prefixes", arXiv e-prints, 2021.

@ARTICLE{2021arXiv211005909W,
       author = {Wick, Christoph and Zöllner, Jochen and Grüning, Tobias},
        title = "{Rescoring Sequence-to-Sequence Models for Text Line Recognition with CTC-Prefixes}",
      journal = {arXiv e-prints},
         year = 2021,
}