Currently we have launch files for our missions which use isaac_ros for efficiency, but for testing on platforms that dont support isaac_ros it would be ideal to have something else: ultralytics. In perception_auv add launch files which should subscribe to image and camera topics from the simulator, pass them to the relevant YOLO node in vortex-deep-learning-pipelines and then run the mission specific code after that (valve-detection, line-fitting etc.)
Currently we have launch files for our missions which use isaac_ros for efficiency, but for testing on platforms that dont support isaac_ros it would be ideal to have something else: ultralytics. In perception_auv add launch files which should subscribe to image and camera topics from the simulator, pass them to the relevant YOLO node in vortex-deep-learning-pipelines and then run the mission specific code after that (valve-detection, line-fitting etc.)