HomeAboutJobsNews
Contact
← Back to Blog

IM Scientists Set New Benchmark for Weakly Supervised Text Recognition (OCR) at CVPR 2020

July 17, 2020

For the first time, IM scientists demonstrated a method ("OrigamiNet") to transcribe challenging handwritten text without line break info. OrigamiNet also achieves state-of-the-art OCR accuracy, even compared to fully supervised methods using line segmentation data.

IM Scientists Set New Benchmark for Weakly Supervised Text Recognition (OCR) at CVPR 2020

Existing OCR techniques for handwritten text transcription typically require large quantities of fully annotated example data to train effectively. These annotations not only require the transcription of the characters themselves, but also manual segmentation of the locations of the line breaks.

‍

For the first time, IM scientists have demonstrated a method ("OrigamiNet") that can transcribe entire paragraphs of challenging handwritten text, for example in old manuscripts or other hard-to-read documents, without having to supply this line break information. After applying the method to multiple benchmark datasets, OrigamiNet also achieves state-of-the-art results on character recognition accuracy, even compared to fully supervised methods that have line segmentation available.

‍
The full paper on this method was published at CVPR 2020, the premier annual computer vision conference, and can be read here: https://arxiv.org/abs/2006.07491

‍

If you are interested in learning how to apply this work to your own business problems, please contact us for more information.

IM Scientists Set New Benchmark for Weakly Supervised Text Recognition (OCR) at CVPR 2020IM Scientists Set New Benchmark for Weakly Supervised Text Recognition (OCR) at CVPR 2020

Read more

IM's hCaptcha Product Suite Now The Largest Independent CAPTCHA Service

IM's hCaptcha Product Suite Now The Largest Independent CAPTCHA Service

Together, hCaptcha and hCaptcha Enterprise now protect hundreds of millions of users across tens of millions of websites and apps every month. Our story shows that you can compete with Big Tech when you put privacy first.
IM Scientists Set New Benchmark for Weakly Supervised Text Recognition (OCR) at CVPR 2020

IM Scientists Set New Benchmark for Weakly Supervised Text Recognition (OCR) at CVPR 2020

For the first time, IM scientists demonstrated a method ("OrigamiNet") to transcribe challenging handwritten text without line break info. OrigamiNet also achieves state-of-the-art OCR accuracy, even compared to fully supervised methods using line segmentation data.
Cloudflare chooses IM's hCaptcha Enterprise offering to protect 12% of the Internet

Cloudflare chooses IM's hCaptcha Enterprise offering to protect 12% of the Internet

Cloudflare (NYSE: NET) today announced that it has chosen IM's hCaptcha Enterprise offering to protect over 25 million customer sites, dropping Google's reCAPTCHA offering.
Meet IM scientists at CVPR 2020 in Seattle

Meet IM scientists at CVPR 2020 in Seattle

IM scientists will be presenting class-leading results in page-level text recognition at CVPR 2020, the top computer vision conference.
IM product featured in FastCompany

IM product featured in FastCompany

One of Intuition Machines’ products for high volume ML annotation, hCaptcha.com, was featured by FastCompany recently.
Meet IM scientists at ICLR 2019 in New Orleans

Meet IM scientists at ICLR 2019 in New Orleans

IM scientists will be presenting recent results in workshops at ICLR 2019.

View all‎‎‎  →‎‎‎

View Blog
HomeAboutJobsNewsContact
©2022 Intuition Machines, Inc. ("IM") All rights reserved. San Francisco, CA, USA
Intuition Machines® is a registered trademark of IM. Not associated with Carlos E. Perez or Intuition Machine Inc.
Privacy PolicyCookie Policy