Enhanced Optical Character Recognition by Optical Sensor Combined with BERT and Cosine Similarity Scoring

Woohyeon Moon; Sarvar Nengroo; Taeyoung Kim; Jihui Lee; Seungah Son; Dongsoo Har

doi:10.1609/aaai.v38i21.30483

Enhanced Optical Character Recognition by Optical Sensor Combined with BERT and Cosine Similarity Scoring

Woohyeon Moon
, Sarvar Nengroo
, Taeyoung Kim
, Jihui Lee
, Seungah Son
, Dongsoo Har

Research output: Contribution to journal › Conference article › peer-review

Abstract

Optical character recognition(OCR) is the technology to identify text characters embedded within images. Conventional OCR models exhibit performance degradation when performing with noisy images. To solve this problem, we propose a novel model, which combines computer vision using optical sensor with natural language processing by bidirectional encoder representations from transformers(BERT) and cosine similarity scoring. The proposed model uses a confidence rate to determine whether to utilize optical sensor alone or BERT/cosine similarity scoring combined with the optical sensor. Experimental results show that the proposed model outperforms approximately 4.34 times better than the conventional OCR.

Original language	English
Pages (from-to)	23585-23586
Number of pages	2
Journal	Proceedings of the AAAI Conference on Artificial Intelligence
Volume	38
Issue number	21
DOIs	https://doi.org/10.1609/aaai.v38i21.30483
State	Published - 25 Mar 2024
Externally published	Yes
Event	38th AAAI Conference on Artificial Intelligence, AAAI 2024 - Vancouver, Canada Duration: 20 Feb 2024 → 27 Feb 2024

Bibliographical note

Publisher Copyright:
Copyright © 2024, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.

ASJC Scopus subject areas

Artificial Intelligence

Access to Document

10.1609/aaai.v38i21.30483

Cite this

@article{3a6ed8b3aea24c6696a029e3a159148c,

title = "Enhanced Optical Character Recognition by Optical Sensor Combined with BERT and Cosine Similarity Scoring",

abstract = "Optical character recognition(OCR) is the technology to identify text characters embedded within images. Conventional OCR models exhibit performance degradation when performing with noisy images. To solve this problem, we propose a novel model, which combines computer vision using optical sensor with natural language processing by bidirectional encoder representations from transformers(BERT) and cosine similarity scoring. The proposed model uses a confidence rate to determine whether to utilize optical sensor alone or BERT/cosine similarity scoring combined with the optical sensor. Experimental results show that the proposed model outperforms approximately 4.34 times better than the conventional OCR.",

author = "Woohyeon Moon and Sarvar Nengroo and Taeyoung Kim and Jihui Lee and Seungah Son and Dongsoo Har",

note = "Publisher Copyright: Copyright {\textcopyright} 2024, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.; 38th AAAI Conference on Artificial Intelligence, AAAI 2024 ; Conference date: 20-02-2024 Through 27-02-2024",

year = "2024",

month = mar,

day = "25",

doi = "10.1609/aaai.v38i21.30483",

language = "English",

volume = "38",

pages = "23585--23586",

journal = "Proceedings of the AAAI Conference on Artificial Intelligence",

issn = "2159-5399",

publisher = "Association for the Advancement of Artificial Intelligence",

number = "21",

}

TY - JOUR

T1 - Enhanced Optical Character Recognition by Optical Sensor Combined with BERT and Cosine Similarity Scoring

AU - Moon, Woohyeon

AU - Nengroo, Sarvar

AU - Kim, Taeyoung

AU - Lee, Jihui

AU - Son, Seungah

AU - Har, Dongsoo

PY - 2024/3/25

Y1 - 2024/3/25

N2 - Optical character recognition(OCR) is the technology to identify text characters embedded within images. Conventional OCR models exhibit performance degradation when performing with noisy images. To solve this problem, we propose a novel model, which combines computer vision using optical sensor with natural language processing by bidirectional encoder representations from transformers(BERT) and cosine similarity scoring. The proposed model uses a confidence rate to determine whether to utilize optical sensor alone or BERT/cosine similarity scoring combined with the optical sensor. Experimental results show that the proposed model outperforms approximately 4.34 times better than the conventional OCR.

AB - Optical character recognition(OCR) is the technology to identify text characters embedded within images. Conventional OCR models exhibit performance degradation when performing with noisy images. To solve this problem, we propose a novel model, which combines computer vision using optical sensor with natural language processing by bidirectional encoder representations from transformers(BERT) and cosine similarity scoring. The proposed model uses a confidence rate to determine whether to utilize optical sensor alone or BERT/cosine similarity scoring combined with the optical sensor. Experimental results show that the proposed model outperforms approximately 4.34 times better than the conventional OCR.

UR - https://www.scopus.com/pages/publications/85189619218

U2 - 10.1609/aaai.v38i21.30483

DO - 10.1609/aaai.v38i21.30483

M3 - Conference article

AN - SCOPUS:85189619218

SN - 2159-5399

VL - 38

SP - 23585

EP - 23586

JO - Proceedings of the AAAI Conference on Artificial Intelligence

JF - Proceedings of the AAAI Conference on Artificial Intelligence

IS - 21

T2 - 38th AAAI Conference on Artificial Intelligence, AAAI 2024

Y2 - 20 February 2024 through 27 February 2024

ER -

Enhanced Optical Character Recognition by Optical Sensor Combined with BERT and Cosine Similarity Scoring

Abstract

Bibliographical note

ASJC Scopus subject areas

Access to Document

Fingerprint

Cite this