Running Autoware without CUDA#
Although CUDA installation is recommended to achieve better performance for object detection and traffic light recognition in Autoware Universe, it is possible to run these algorithms without CUDA. The following subsections briefly explain how to run each algorithm in such an environment.
Running 2D/3D object detection without CUDA#
Autoware Universe's object detection can be run using one of five possible configurations:
Of these five configurations, only the last one (
euclidean_cluster) can be run without CUDA. For more details, refer to the
euclidean_cluster module's README file.
Running traffic light detection without CUDA#
For traffic light recognition (both detection and classification), there are two modules that require CUDA:
To run traffic light detection without CUDA, set
false in the traffic light launch file. Doing so disables the
traffic_light_ssd_fine_detector such that traffic light detection is handled by the
map_based_traffic_light_detector module instead.
To run traffic light classification without CUDA, set
false in the traffic light classifier launch file. Doing so will force the
traffic_light_classifier to use a different classification algorithm that does not require CUDA or a GPU.