EfficientOCR: An Extensible, Open-Source Package for Efficiently Digitizing World Knowledge

Tom Bryan, Jacob Carlson, Abhishek Arora, Melissa Dell


Abstract
Billions of public domain documents remain trapped in hard copy or lack an accurate digitization. Modern natural language processing methods cannot be used to index, retrieve, and summarize their texts; conduct computational textual analyses; or extract information for statistical analyses, and these texts cannot be incorporated into language model training. Given the diversity and sheer quantity of public domain texts, liberating them at scale requires optical character recognition (OCR) that is accurate, extremely cheap to deploy, and sample-efficient to customize to novel collections, languages, and character sets. Existing OCR engines, largely designed for small-scale commercial applications in high resource languages, often fall short of these requirements. EffOCR (EfficientOCR), a novel open-source OCR package, meets both the computational and sample efficiency requirements for liberating texts at scale by abandoning the sequence-to-sequence architecture typically used for OCR, which takes representations from a learned vision model as inputs to a learned language model. Instead, EffOCR models OCR as a character or word-level image retrieval problem. EffOCR is cheap and sample efficient to train, as the model only needs to learn characters’ visual appearance and not how they are used in sequence to form language. Models in the EffOCR model zoo can be deployed off-the-shelf with only a few lines of code and include lightweight models designed for mobile phones that are extremely cheap to deploy. Importantly, EffOCR also allows for easy, sample efficient customization with a simple model training interface and minimal labeling requirements due to its sample efficiency. We illustrate the utility of EffOCR by cheaply and accurately digitizing 20 million historical U.S. newspaper scans, evaluating zero-shot performance on randomly selected documents from the U.S. National Archives, and accurately digitizing a Japanese document collection for which all other OCR solutions failed.
Anthology ID:
2023.emnlp-demo.52
Volume:
Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing: System Demonstrations
Month:
December
Year:
2023
Address:
Singapore
Editors:
Yansong Feng, Els Lefever
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
579–596
Language:
URL:
https://aclanthology.org/2023.emnlp-demo.52
DOI:
10.18653/v1/2023.emnlp-demo.52
Bibkey:
Cite (ACL):
Tom Bryan, Jacob Carlson, Abhishek Arora, and Melissa Dell. 2023. EfficientOCR: An Extensible, Open-Source Package for Efficiently Digitizing World Knowledge. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, pages 579–596, Singapore. Association for Computational Linguistics.
Cite (Informal):
EfficientOCR: An Extensible, Open-Source Package for Efficiently Digitizing World Knowledge (Bryan et al., EMNLP 2023)
Copy Citation:
PDF:
https://aclanthology.org/2023.emnlp-demo.52.pdf
Video:
 https://aclanthology.org/2023.emnlp-demo.52.mp4