OCR Post-processing Using Weighted Finite-State Transducers

A new approach for Stochastic Error-Correcting Language Modeling based on Weighted Finite-State Transducers (WFSTs) is proposed as a method to post-process the results of an optical character recognizer (OCR). Instead of using the recognized string as an input to the transducer, in our approach the complete set of OCR hypotheses, a sequence of vectors of a posteriori class probabilities, is used to build a WFST that is then composed with independent WFSTs for the error and language models. This combines the practical advantages of a de-coupled (OCR + post-processor) model with the full power of an integrated model.