A new dataset and open-source OCR pipeline transcribes medieval English legal manuscripts at up to 88% word accuracy using CNN+LSTM and language model correction.
IEEE Trans
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Democratizing the medieval English legal tradition
A new dataset and open-source OCR pipeline transcribes medieval English legal manuscripts at up to 88% word accuracy using CNN+LSTM and language model correction.