How Computers Use Neural Networks to Read Messy Handwriting
A system that uses artificial intelligence to identify text in images by combining dictionary lookups with neural networks that analyze visual traits like handwriting style or blur.
Original patent title: “System and method of character recognition using fully convolutional neural networks with attention”
A system that uses artificial intelligence to identify text in images by combining dictionary lookups with neural networks that analyze visual traits like handwriting style or blur. Granted to Kodak Alaris Inc in 2023 with 7 claims.
Key facts
Coverage
What does this patent actually cover?
This system improves Optical Character Recognition (OCR) by breaking down images of text into individual words. It first checks if a word matches a known dictionary entry to provide a quick, high-confidence prediction. If the word is not in the dictionary, the system assigns 'qualitative descriptors'—like the slant of the letters, the amount of blur, or the type of paper—to help a second neural network better understand the messy or unusual text. By using these descriptors as a guide, the system can perform a more accurate probabilistic correction to guess what the word actually says.
The gap
What does this patent NOT cover?
- Does not cover basic OCR that relies solely on template matching without neural network descriptors.
- Does not cover systems that process text without first segmenting the image into word blocks based on whitespace.
- Does not cover audio-to-text transcription systems.
- Does not cover methods that do not use a dictionary-based verification step.
These exclusions are unique to PatentBrief — derived from the actual claim language, not patent-office boilerplate.
What made this novel
Instead of just trying to read the letters, the system uses a 'steering factor'—essentially an attention mechanism—that tells the neural network which parts of the image are most important based on visual traits like skew or blur.
Schematic visualization of the patent's claim structure. Hand-drawn diagrams in progress for each landmark patent.
Where you've seen this
Real-world examples
Automated document processing for insurance claimsclaimsThe numbered statements at the end of a patent that legally define what the inventor owns.Read more →
Digitization of handwritten historical archives
Smart scanning software for business invoices
Why it matters
The bigger picture
Traditional OCR often struggles with handwritten notes, faded documents, or poor-quality scans. By using neural networks to 'describe' the visual quality of the text before trying to read it, this technology helps digitize historical archives or complex business forms that were previously too difficult for computers to interpret accurately.
Filed
October 20, 2020
Granted
August 1, 2023
Market context
Who's building on this
Companies in this space
Kodak Alaris continues to specialize in information management and document imaging solutions. The technology described here is part of a broader trend where companies like Google and Microsoft integrate attention-based neural networks into their document cloud services to handle non-standardized text.
Market impact
This patent represents an incremental but vital improvement in the field of document capture. It addresses the 'last mile' problem in OCR, where high-quality digital conversion is required for non-uniform, real-world documents, thereby reducing the need for manual human data entry in enterprise workflows.
Claim 1 — Plain English
What this patent covers
This system improves Optical Character Recognition (OCR) by breaking down images of text into individual words. It first checks if a word matches a known dictionary entry to provide a quick, high-confidence prediction. If the word is not in the dictionary, the system assigns 'qualitative descriptors'—like the slant of the letters, the amount of blur, or the type of paper—to help a second neural network better understand the messy or unusual text. By using these descriptors as a guide, the system can perform a more accurate probabilistic correction to guess what the word actually says.
The clever bit
Instead of just trying to read the letters, the system uses a 'steering factor'—essentially an attention mechanism—that tells the neural network which parts of the image are most important based on visual traits like skew or blur.
What it does not cover
- Does not cover basic OCR that relies solely on template matching without neural network descriptors.
- Does not cover systems that process text without first segmenting the image into word blocks based on whitespace.
- Does not cover audio-to-text transcription systems.
- Does not cover methods that do not use a dictionary-based verification step.
Patent timeline
Application submitted to the patent office
Application published, typically 18 months after filing
Patent officially issued
PatentBrief Score
Impact Score
Early stage
Citation count
0/40
No citations yet
Claim breadth
5/20
Moderate scope
Recency
20/20
Granted within 5 years
Assignee scale
0/20
Independent or smaller assigneeassigneeThe entity that owns the patent — usually the inventor's employer or a company.Read more →
PatentBrief Impact Score — based on citation count, claim breadth, recency, and assignee scale. Not a legal assessment.
Heuristic Value Estimate
What this patent might be worth
$24K – $77K
Midpoint $48K · 14.3 yr remaining · industry ×1.6
Heuristic only — blends forward/backward citation counts, claim scope, time remaining, litigation history, and CPC-derived industry baseline. Real valuations need a professional appraisal.
The original legal language
Original claims
7 claims as filed with the patent office.
Concepts involved
Citations
Patent lineage
Cite this patent
SUCH, F. P., BROCKLER, F., Ptucha, R., & HUTKOWSKI, P. (2023). How Computers Use Neural Networks to Read Messy Handwriting (U.S. Patent No. 11,715,014). U.S. Patent and Trademark Office. https://patentbrief.org/patent/us/11715014/gemini-multimodal-model
Auto-generated from the patent record. Double-check author order and the issue date against the official USPTO document before submitting.
Embed
Add this patent to your site
Drop this plain-English patent card into any blog post or article — free, no signup. It always links back to the full breakdown here.
<div data-patentlens-widget data-patent-number="US11715014"></div> <script src="https://patentbrief.org/embed.js" async></script>
Stay in the loop
Get a weekly digest of new patents.
One email per week. No spam. Unsubscribe anytime.
Keep exploring
Related patents you should know
US 4683195 · 1987
How to Make Billions of Copies of a DNA Segment
This patent describes the Polymerase Chain Reaction (PCR), a method to rapidly create many copies of a specific piece of DNA or RNA, enabling its detection and analysis.
Cetus Corp
US 8697359 · 2014
How to Edit Genes in Human Cells Using an Engineered CRISPR System
This patent describes an engineered CRISPR-Cas9 system for precisely cutting DNA in eukaryotic cells to change how genes work, opening the door for gene editing in complex organisms.
Massachusetts Institute of Technology
US 7657849 · 2010
How the iPhone's Slide-to-Unlock Gesture Works
Apple's 2010 patent describes unlocking a device by dragging a specific graphical image across the touchscreen along a predefined path, a gesture that became iconic with the original iPhone.
Apple Inc
US 4733665 · 1988
How Doctors Implant a Permanent Stent Using a Balloon
This patent describes the method for placing a permanent, expandable wire mesh tube inside a blood vessel or other body tube using a balloon-tipped catheter to widen it and keep it open.
Expandable Grafts Partnership
US 4965188 · 1990
How to Make Many Copies of a DNA Piece with Heat
This patent describes the Polymerase Chain Reaction (PCR) method, a technique to make millions of copies of a specific DNA segment using a heat-resistant enzyme and repeated temperature changes.
Cetus Corp
US 4235871 · 1980
How to Encapsulate Active Materials in Lipid Bubbles Efficiently
This patent describes a method for trapping biologically active substances inside tiny, multi-layered fat bubbles called liposomes, using a specific water-in-oil emulsion and gel-forming process to improve how much material gets captured.
Individual
More to explore
More in AI & Machine Learning
US 10452978 · 2019 · Google LLC
How AI Models Understand Language Using 'Attention'
US 6523026 · 2003 · Huntsman International LLC
How Computers Find Hidden Connections Between Different Fields of Knowledge
US 11615208 · 2023 · Capital One Services LLC
How Cloud Systems Automatically Create and Train AI Data Models
US 10402750 · 2019 · Facebook Inc
How Facebook Uses Deep Learning to Predict What You Might Like
New to patents?
Common Questions
Frequently Asked Questions
What does How Computers Use Neural Networks to Read Messy Handwriting cover?
A system that uses artificial intelligence to identify text in images by combining dictionary lookups with neural networks that analyze visual traits like handwriting style or blur.
Who owns patent US 11715014?
Kodak Alaris Inc owns this patent, granted in 2023.
When does this patent expire?
This patent is expected to expire on August 1, 2043, when the invention enters the public domain.
What problem does this patent solve?
Traditional OCR often struggles with handwritten notes, faded documents, or poor-quality scans. By using neural networks to 'describe' the visual quality of the text before trying to read it, this technology helps digitize historical archives or complex business forms that were previously too difficult for computers to interpret accurately.
What does this patent NOT cover?
Does not cover basic OCR that relies solely on template matching without neural network descriptors.
Patent monitoring