Generating scene graphs from digital images using external knowledge and image reconstruction

This invention describes a computer system that analyzes digital images to identify objects and their relationships. It uses information from a vast external knowledge source to refine its understanding of these objects and their connections, then creates a structured map (a scene graph) of what's happening in the image. The system can also learn and improve by generating images based on its understanding.

Why it matters: Filed before the widespread availability of powerful Vision-Language Models. These models can now provide richer external knowledge and more accurate object/relationship understanding, making the system's core function more robust and easier to implement.

Status: Not in forceListed as no longer active. The specific reason is not in the record we hold.
How hard to build: SpecializedSoftware logic, machine learning, external knowledgebase integration

3 AI reinvention directions for this patent

See all three

AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.

Reinvent this with AI

Where this one came from

Every morning we go through millions of abandoned patents and pick out 15 worth your time. This was one of today’s.

This one

+10