CARLA
An open-source autonomous driving simulation platform that provides urban driving scenarios and sensor simulation, widely used for research and testing of autonomous driving algorithms.
Timeline (last 90 days)
Jia Xiaosong used ThinkTwice to run a large number of closed-loop experiments in the CARLA simulator, collecting nearly unlimited data and investing 100TB of storage and substantial compute to train larger models, but driving capability did not jump from "drives okay" to "drives perfectly" as data and model scale grew.
In autonomous driving, synthetic data is typically produced by game engines like Unreal Engine and Unity, simulation platforms like CARLA and NVIDIA DRIVE Sim, or generative models like diffusion models and NeRF
All Facts (2)
In autonomous driving, synthetic data is typically produced by game engines like Unreal Engine and Unity, simulation platforms like CARLA and NVIDIA DRIVE Sim, or generative models like diffusion models and NeRF
90%UnverifiedJia Xiaosong used ThinkTwice to run a large number of closed-loop experiments in the CARLA simulator, collecting nearly unlimited data and investing 100TB of storage and substantial compute to train larger models, but driving capability did not jump from "drives okay" to "drives perfectly" as data and model scale grew.
50%