Systems and methods for graph-based AI training

This invention describes a software system that trains artificial intelligence by first converting real-world scenarios into a knowledge graph. This graph specifically represents objects, their locations, movements over time, and their meanings. The AI is then trained by navigating this structured graph, which helps it learn more effectively.

Why it matters: Filed in 2022, the rapid advancements in graph neural networks and the increasing demand for robust AI in autonomous systems (as indicated by the classification) make this graph-based training approach more relevant and powerful today. The ability to encode complex spatiotemporal and semantic data into graphs addresses key challenges in developing reliable AI for real-world applications.

Status: Not in forceListed as no longer active. The specific reason is not in the record we hold.
How hard to build: SpecializedComplex AI software, data processing, algorithms

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