search

PRBonn / lidar-bonnetal

Semantic and Instance Segmentation of LiDAR point clouds for autonomous driving
http://semantic-kitti.org
MIT License
961 stars 206 forks source link

How to adapt 16 beam lidar? #55

Closed lihao2333 closed 3 years ago

lihao2333 commented 4 years ago

My lidar has only 16 beams.
How can I get the best result based on your work?
I want to change train the model with the resolution of 2048 * 16, could this help the result?

BenjaminYoung29 commented 4 years ago

@lihao2333 Hi, have you done the job? My idea is to downsample the original dataset from 204864 to 204816 and train on it. Then use it on the VDL16 LiDAR. How do you think of it?

jbehley commented 4 years ago

Hi @BenjaminYoung29,

please see my answer in https://github.com/PRBonn/lidar-bonnetal/issues/69#issuecomment-711153879.

hope that helps.

BenjaminYoung29 commented 4 years ago

Hi @BenjaminYoung29,

please see my answer in #69 (comment).

hope that helps.

Hi @jbehley

thanks for your reply.

I will dig into the laserscan.py to see how the configuration works.

Thanks again.

jbehley commented 3 years ago

Did it solve your problem? If no, please reopen the issue.