AI That Finds Partially Hidden Objects in Images
This patent describes an AI system that uses multiple machine learning models to accurately identify objects in an image, even when parts of those objects are blocked by something else.
Original patent title: “Systems, methods, and computer programs for using a network of machine learning models to detect an image depicting an object of interest which can be partially occluded by another object”
This patent describes an AI system that uses multiple machine learning models to accurately identify objects in an image, even when parts of those objects are blocked by something else. Granted in 2026.
Coverage
What does this patent actually cover?
This patent details a system for detecting objects in images, even when they are partially hidden. It starts by taking an input image and extracting 'object-based features' and 'context-based features'. These features are then processed by a deep convolutional neural network (DCNN) model. The DCNN's output is sent to two specialized models: an 'occlusion model' and a 'generative compositional model'. The occlusion model specifically identifies areas where an object is blocked by another object. The generative compositional model looks for different parts or components of objects that might be present. By combining the information from both these models, the system creates a 'likelihood map' to show if a particular object is depicted in the image, even with parts missing. For example, if a car is mostly hidden behind a tree, this system could still identify it by recognizing visible parts and understanding the context of the occlusion.
The gap
What does this patent NOT cover?
- Does not cover detecting objects that are completely hidden and have no visible parts or context clues.
- Does not cover object detection methods that do not use a deep convolutional neural network (DCNN).
- Does not cover systems that only use a single model for object detection without separate occlusion and compositional analysis.
- Does not cover general image classification where the goal is to label the entire image, rather than locate specific objects.
- Does not cover object detection that relies solely on traditional computer vision techniques without machine learning models.
These exclusions are unique to PatentBrief — derived from the actual claim language, not patent-office boilerplate.
Key facts
What made this novel
The clever part is combining an 'occlusion model' that identifies blocked areas with a 'generative compositional model' that recognizes object parts. This dual approach allows the system to piece together what an object is, even when it's not fully visible, by understanding both what's missing and what's present.
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
Autonomous vehicle perception systems identifying pedestrians or other cars partially blocked by obstacles.
Security camera systems detecting intruders partially hidden behind furniture or foliage.
Industrial automation robots identifying components on an assembly line that are partially stacked or covered.
Medical imaging analysis for detecting anomalies partially obscured by other tissues.
Why it matters
The bigger picture
Accurately detecting objects, even when they are partially obscured, is critical for many real-world applications. This technology improves the reliability of systems that need to 'see' and understand their environment, like self-driving cars or security cameras. It helps these systems make more informed decisions by reducing errors caused by visual obstructions.
Filed
June 14, 2022
Granted
September 15, 2026
Market context
Who's building on this
Companies in this space
Major technology companies like Google, Meta, and Amazon, along with specialized AI startups, are continuously advancing object detection and computer vision. Companies developing autonomous driving technology, such as Waymo and Cruise, are particularly interested in robust object detection under occlusion. Manufacturers of smart cameras and security systems also integrate similar capabilities into their products.
Market impact
Improved object detection, especially with occlusion handling, significantly enhances the performance and safety of AI-driven systems. This capability is essential for the widespread adoption of autonomous vehicles and advanced robotics, as it allows them to operate more reliably in complex, real-world environments. It also drives innovation in surveillance, quality control, and augmented reality applications, making these technologies more effective and trustworthy.
Claim 1 — Plain English
What this patent covers
This patent details a system for detecting objects in images, even when they are partially hidden. It starts by taking an input image and extracting 'object-based features' and 'context-based features'. These features are then processed by a deep convolutional neural network (DCNN) model. The DCNN's output is sent to two specialized models: an 'occlusion model' and a 'generative compositional model'. The occlusion model specifically identifies areas where an object is blocked by another object. The generative compositional model looks for different parts or components of objects that might be present. By combining the information from both these models, the system creates a 'likelihood map' to show if a particular object is depicted in the image, even with parts missing. For example, if a car is mostly hidden behind a tree, this system could still identify it by recognizing visible parts and understanding the context of the occlusion.
The clever bit
The clever part is combining an 'occlusion model' that identifies blocked areas with a 'generative compositional model' that recognizes object parts. This dual approach allows the system to piece together what an object is, even when it's not fully visible, by understanding both what's missing and what's present.
What it does not cover
- Does not cover detecting objects that are completely hidden and have no visible parts or context clues.
- Does not cover object detection methods that do not use a deep convolutional neural network (DCNN).
- Does not cover systems that only use a single model for object detection without separate occlusion and compositional analysis.
- Does not cover general image classification where the goal is to label the entire image, rather than locate specific objects.
- Does not cover object detection that relies solely on traditional computer vision techniques without machine learning models.
Patent timeline
Application submitted to the patent office
Patent officially issued
PatentBrief Score
Impact Score
Early stage
Citation count
0/40
No citations yet
Claim breadth
0/20
Narrow claimsclaimsThe numbered statements at the end of a patent that legally define what the inventor owns.Read more →
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
$19K – $61K
Midpoint $38K · 15.7 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.
Claim text not yet imported for this patent
Concepts involved
Cite this patent
(2026). AI That Finds Partially Hidden Objects in Images (U.S. Patent No. 12,738,026). U.S. Patent and Trademark Office. https://patentbrief.org/patent/us/12738026/systems-methods-and-computer-programs-for-using-a-network-of-machine-learning
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="US12738026"></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
Semantically similar
You might also find these interesting
US 11023780 · 2021 · Superb AI Co Ltd
How AI Automatically Labels Images and Checks Its Own Work
US 11715014 · 2023 · Kodak Alaris Inc
How Computers Use Neural Networks to Read Messy Handwriting
US 20230162023 · Mitsubishi Electric Research Laboratories
Automated AI for Adapting to New Data Without Retraining
US 8862582 · 2014 · AT&T Intellectual Property I LP
How Computers Automatically Organize and Search Photos Using Contextual Data
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 AI That Finds Partially Hidden Objects in Images cover?
This patent describes an AI system that uses multiple machine learning models to accurately identify objects in an image, even when parts of those objects are blocked by something else.
When does this patent expire?
This patent is expected to expire on September 15, 2046, when the invention enters the public domain.
What problem does this patent solve?
Accurately detecting objects, even when they are partially obscured, is critical for many real-world applications. This technology improves the reliability of systems that need to 'see' and understand their environment, like self-driving cars or security cameras. It helps these systems make more informed decisions by reducing errors caused by visual obstructions.
What does this patent NOT cover?
Does not cover detecting objects that are completely hidden and have no visible parts or context clues.
Patent monitoring