Systems and methods for detection and correction of ocr text

Patent metadata

Field	Value
Publication number	US-2023083000-A1
Application number	US-202217895818-A
Country	US
Kind code	A1
Filing date	Aug 25, 2022
Priority date	Aug 27, 2021
Publication date	Mar 16, 2023
Grant date	—

Abstract

OCR-text correction system and method embodiments are described. The OCR-text correction embodiments comprise or cooperate with a transformer-based sequence-to-sequence language model. The model is pretrained to denoise corrupted text and is fine-tuned using OCR-correction-specific examples. Text obtained at least in part through OCR is applied to the fine-tuned pretrained transformer model to detect at least one error in a subset of the text. Responsive to detecting the at least one error, the fine-tuned pretrained transformer model outputs an updated subset of the text to correct the at least one error.

First claim

What is claimed is:

1 . A computer-implemented method, comprising:
accessing a machine learning model that is pretrained by a first training dataset, the machine learning model pretrained to perform a non-optical character recognition (non-OCR) task; adjusting the machine learning model using a second training dataset, the second training dataset comprising OCR samples, the machine learning model adjusted to perform an OCR task; receiving a document that includes text obtained at least in part through OCR; applying the adjusted machine learning model to the text to detect at least one error in a subset of the text; and outputting an updated subset of the text to correct the at least one error in the subset of the text.

2 . The method of claim 1 , wherein the pretrained transformer model is bidirectional autoregressive transformer model, the bidirectional autoregressive transformer model including:
a bidirectional encoder configured to receive the text; and an autoregressive decoder configured to detect the at least one error in the text and correct the at least one error in the text by predicting original text.

3 . The method of claim 1 , wherein the first training dataset includes one or more of the following: token masking, token deletion, sentence permutation, document rotation, or text infilling.

4 . The method of any of claim 1 , wherein the second training dataset includes monograph and periodical example sentences.

5 . The method of claim 1 , wherein the fine-tuned pretrained transformer model is configured to perform the detection and correction of the at least one error in a single step.

6 . The method of claim 1 , wherein the transformer model is configured to correct the at least one error in the text without being trained on alignment characters.

7 . The method of claim 1 , wherein the first training dataset comprises fewer than 1,000 documents.

8 . The method of claim 1 , wherein the at least one error includes an oversegmentation error caused by incorrectly segmenting a single word into two separate words by OCR.

9 . The method of claim 1 , wherein the at least one error includes an undersegmentation error caused by incorrectly combining a plurality of words into a single word by OCR.

10 . The method of claim 1 , wherein the at least one error includes a misrecognized character error caused by incorrectly recognizing a character by OCR.

11 . The method of claim 1 , wherein the at least one error includes a missing character error caused by incorrectly omitting a character by OCR.

12 . The method of claim 1 , wherein the at least one error includes a hallucination error caused by incorrectly inserting a non-existing character by OCR.

13 . A computer system for detecting and/or correcting text, comprising:
a processor; and memory in communication with the processor, the memory configured to store instructions that, when executed by the processor, cause the processor to:
access a pretrained transformer model pretrained using a first training dataset;
fine-tune the pretrained transformer model using a second training dataset;
provide text obtained at least in part through optical character recognition (OCR);
apply the text to the fine-tuned pretrained transformer model to detect at least one error in a subset of the text; and
output an updated subset of the text by the fine-tuned pretrained transformer model to correct the at least one error in the subset of the text.

14 . The computer system of claim 13 , wherein the pretrained transformer model is bidirectional autoregressive transformer model including:
a bidirectional encoder configured to receive the text; and an autoregressive decoder configured to detect the at least one error in the text and correct the at least one error in the text by predicting original text.

15 . The computer system of claim 13 wherein the first training dataset includes one or more of the following: token masking, token deletion, sentence permutation, document rotation, or text infilling.

16 . The computer system of claim 13 , wherein the second training dataset includes monograph and periodical example sentences.

17 . The computer system of claim 13 , wherein the fine-tuned pretrained transformer model is configured to perform the detection and correction of the at least one error in a single step.

18 . The computer system of claim 13 , wherein the transformer model is configured to correct the at least one error in the text without being trained on alignment characters.

19 . The computer system of claim 13 , wherein the first training dataset comprises fewer than 1,000 documents.

20 . A non-transitory computer readable storage medium configured to store code comprising instructions, wherein the instructions, when executed by a processor, cause the processor to:
access a pretrained transformer model pretrained using a first training dataset; fine-tune the pretrained transformer model using a second training dataset; provide text obtained at least in part through optical character recognition (OCR); apply the text to the fine-tuned pretrained transformer model to detect at least one error in a subset of the text; and output an updated subset of the text by the fine-tuned pretrained transformer model to correct the at least one error in the subset of the text.

Assignees

• Ancestry Com Operations Inc

Inventors

• Fujimoto Masaki Stanley
• Yu Yen-Yun

Classifications

• G06V30/26

  Techniques for post-processing, e.g. correcting the recognition result · CPC title

• G06V30/19147

  Obtaining sets of training patterns; Bootstrap methods, e.g. bagging or boosting · CPC title

• G06V30/133 Primary

  Evaluation of quality of the acquired characters · CPC title