search

unmannedlab / RELLIS-3D

RELLIS-3D: A Multi-modal Dataset for Off-Road Robotics
https://unmannedlab.github.io/research/RELLIS-3D
276 stars 48 forks source link

Integration with ROS #5

Open ArghyaChatterjee opened 3 years ago

ArghyaChatterjee commented 3 years ago

Hello, I was wondering if the whole setup and prediction can be done online using ROS. I mean once we train a model (benchmark models provided by you) using Ground truth dataset (created by us), can we subscribe to images and pointclouds published in separate topics and predict corresponding semantic labels & publish them in a node as a topic/topics & visualize in rviz ?

If you have done this, please let me know but if you haven't, can you refer to any work/repository which have already implemented this ?

maskjp commented 3 years ago

Hi, @ArghyaChatterjee! Thank you for your interest in RELLIS-3D. We haven't combined the semantic inference with the ros. But there are some good repositories that might be helpful:

https://github.com/ethz-asl/deeplab_ros https://github.com/PRBonn/rangenet_lib https://github.com/LidarPerception/segmenters_lib