PRBonn / lidar-bonnetal

Semantic and Instance Segmentation of LiDAR point clouds for autonomous driving
http://semantic-kitti.org
MIT License
915 stars 204 forks source link

trained models? #70

Closed he-guo closed 3 years ago

he-guo commented 3 years ago

Oh ! My GPU is too old . Do you have any models that have been trained? I want to tag my data directly with this model.

he-guo commented 3 years ago

and can I try this work without GPU and cuda?

jbehley commented 3 years ago

Here you can find pre-trained models there: https://github.com/PRBonn/lidar-bonnetal#semantickitti

But these are trained with a 64 beam Velodyne and usually, these do not transfer well to other sensors. And to run this on the CPU, you might have to modify the code to correctly use the CPU instead of GPU. (Never tried...)

he-guo commented 3 years ago

I see your suggestion. Thanks a lot.

jbehley commented 3 years ago

I think this issue is solved. If you have still questions, just re-open the issue.