This invention describes a computer system that predicts what a car's driver will do next. It works by looking at images from the car's perspective, identifying moving objects and features of the road scene. It then combines this information with what it understands about the driver's intentions to forecast their actions. The system specifically uses neural networks to process the images and identify objects.
Why it matters: Filed when AI for complex scene understanding and intent prediction was still nascent. Advances in deep learning and causal AI since 2020 make building robust intent-aware models more feasible.
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