From 96da0aebbb831d6e8ad1ce274860d63821d47e9e Mon Sep 17 00:00:00 2001 From: kminoda Date: Mon, 30 May 2022 09:53:56 +0900 Subject: [PATCH] added files Signed-off-by: kminoda --- docs/how-to-guides/.pages | 1 + docs/how-to-guides/index.md | 2 +- .../running-autoware-without-cuda.md | 34 +++++++++++++++++++ 3 files changed, 36 insertions(+), 1 deletion(-) create mode 100644 docs/how-to-guides/running-autoware-without-cuda.md diff --git a/docs/how-to-guides/.pages b/docs/how-to-guides/.pages index 003ffecd69b..a1be851925f 100644 --- a/docs/how-to-guides/.pages +++ b/docs/how-to-guides/.pages @@ -1,4 +1,5 @@ nav: - index.md + - running-autoware-without-cuda.md - advanced-usage-of-colcon.md - how-to-integrate-autoware-with-your-vehicle.md diff --git a/docs/how-to-guides/index.md b/docs/how-to-guides/index.md index 36fbba7f3f7..c5aaf5e848a 100644 --- a/docs/how-to-guides/index.md +++ b/docs/how-to-guides/index.md @@ -1,5 +1,5 @@ # How-to guides - +- [Running Autoware without CUDA](running-autoware-without-cuda.md) - [Advanced usage of colcon](advanced-usage-of-colcon.md) - [Integrate Autoware with your vehicle](how-to-integrate-autoware-with-your-vehicle.md) diff --git a/docs/how-to-guides/running-autoware-without-cuda.md b/docs/how-to-guides/running-autoware-without-cuda.md new file mode 100644 index 00000000000..5a3ef8cd597 --- /dev/null +++ b/docs/how-to-guides/running-autoware-without-cuda.md @@ -0,0 +1,34 @@ +# Running Autoware without CUDA + +Two following algorithms in Autoware Universe depend on a CUDA environment. + +- 2D/3D object detection +- traffic light detection + +However, it is still possible to run Autoware in the environment without CUDA. +The following subsections briefly explain how to run these algorithms in such an environment. + +!!! warning + + It is highly recommended to prepare a CUDA environment to achieve better performance. + +## Running 2D/3D object detection without CUDA + +As for the object detection system in Autoware, the following three packages require the CUDA environment: + +- `lidar_centerpoint` +- `lidar_apollo_instance_segmentation` +- `tensorrt_yolo` + +Instead of these modules, you can use the `euclidean_cluster` module, which is a 3D object detection algorithm that does not require a CUDA environment (refer to [the readme](https://github.com/autowarefoundation/autoware.universe/tree/main/perception/euclidean_cluster) for detail). + +## Running traffic light detection without CUDA + +As for the traffic light recognition system in Autoware, there are two package that require the CUDA environment: + +- `traffic_light_ssd_fine_detector` +- `traffic_light_classifier` + +To run traffic light detection without CUDA, set [`enable_fine_detection` in this file](https://github.com/autowarefoundation/autoware.universe/blob/9445f3a7acd645d12a64507c3d3bfa57e74a3634/launch/tier4_perception_launch/launch/traffic_light_recognition/traffic_light.launch.xml#L3) to `false` so that `traffic_light_ssd_fine_detector` be disabled and the traffic light detection solely be executed by `map_based_traffic_light_detector`. + +To run traffic light classification without CUDA, set [`use_gpu` in this file](https://github.com/autowarefoundation/autoware.universe/blob/9445f3a7acd645d12a64507c3d3bfa57e74a3634/perception/traffic_light_classifier/launch/traffic_light_classifier.launch.xml#L7) to `false` so that `traffic_light_classifier` use classification algorithm that does not use CUDA environment or GPU. \ No newline at end of file