Temporal-based perception for autonomous systems and applications

This invention describes a system for autonomous vehicles that uses machine learning to understand its surroundings by combining information from various sensors over time. It first processes images from different sensors into a "fused feature map" for a specific moment. Then, it combines this current map with previously generated fused maps, which are adjusted for temporal changes, to create a comprehensive understanding of the environment. This combined, time-aware information helps the vehicle make decisions.

Why it matters: Filed before the widespread adoption of advanced multi-modal foundation models. The ability of machine learning models to intrinsically fuse diverse sensor data and perform sophisticated temporal transformations has significantly matured since 2023.

Status: Not in forceListed as no longer active. The specific reason is not in the record we hold.
How hard to build: SpecializedComplex machine learning software

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