Perception software interprets vehicle camera feeds to identify roads, lanes, traffic signals and surrounding objects for automated driving systems.

Generative simulation model transforms recorded driving scenes across cameras, varying weather, lighting and objects to expand training and validation data.

Generative world model simulates coordinated camera, lidar and other sensor representations of driving scenes for autonomous system development.

Vision based autonomy software combines perception and driving decisions for automakers developing advanced driver assistance and autonomous driving systems.

Generative video model creates synthetic driving scenes from scratch across multiple cameras to support autonomous driving training and validation.