search

edwardzhou130 / PolarSeg

Implementation for PolarNet: An Improved Grid Representation for Online LiDAR Point Clouds Semantic Segmentation (CVPR 2020)
BSD 3-Clause "New" or "Revised" License
373 stars 80 forks source link

realust visualization #24

Closed yangchandler closed 3 years ago

yangchandler commented 3 years ago

Hi,thx for your resource code. I have one question, I used test_pretrain.py to get infer labels,but when I use SemanticKitti-API to visualize the labels, The visualization results are dark,as if the label does not correspond to the official color.Did you have a question like this?

edwardzhou130 commented 3 years ago

I did not transfer the label back to the original label format which is used in semantic-kitti-api. So the label probably don't match with the colormap https://github.com/edwardzhou130/PolarSeg/blob/3f95ad6ad97889ae2a233fa774deaf77ec77eb90/semantic-kitti.yaml#L37-L71 You can use remap_semantic_labels.py(with --inverse) from semantic-kitti-api to do the inverse transferring.

yangchandler commented 3 years ago

My question has been solved. Thank you for your answer.