Hello ROS community,

I am looking for implementations of 3D object recognition/detection (like cars or people) inside a 3D pointclound obtain using Lidar 3D. I have found methods based on CNN but no ROS wrapper. Does someone is using such algorithm ? Thank you !