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

Train Darknet on data similar to Semantic3D #89

Closed iphyjon closed 2 years ago

iphyjon commented 2 years ago

Dear @jbehley,

Thank you for this work.

So, I want to train the darknet arch with data similar to that of Semantic3D. I know the SemanticKITTI format is different from that of the Semantic3D format. I was wondering if there is a way to convert from Semantic3D to SemanticKITTI data format. Since my data is not in sequential format, the results are poor and I know it has to do with the data format.

I have already change the fov to that of the HDL32e.

Is it possible to generate sequential data from the Semantic3D format to that of the SemanticKITTI format? I don't have the pose.txt or calibration file.

Thanks in advance