Confidence Prediction for Lexicon-Free OCR

TL;DR

This paper introduces two confidence prediction techniques for lexicon-free OCR systems, significantly reducing misreads and improving reliability in real-world applications where explicit confidence scores are essential.

Contribution

The paper presents novel confidence measurement methods tailored for lexicon-free OCR, addressing the challenge of error filtering without lexicon assistance.

Findings

Significant reduction in OCR misreads using proposed confidence methods

Effective confidence prediction demonstrated on standard benchmarks

Proprietary dataset results confirm practical applicability

Abstract

Having a reliable accuracy score is crucial for real world applications of OCR, since such systems are judged by the number of false readings. Lexicon-based OCR systems, which deal with what is essentially a multi-class classification problem, often employ methods explicitly taking into account the lexicon, in order to improve accuracy. However, in lexicon-free scenarios, filtering errors requires an explicit confidence calculation. In this work we show two explicit confidence measurement techniques, and show that they are able to achieve a significant reduction in misreads on both standard benchmarks and a proprietary dataset.