Results on SemanticKITTI test set

fengluodb / RangeSeg

RangeSeg: Efficient Lidar Semantic Segmentation on Range view

MIT License

30 stars 5 forks source link

Results on SemanticKITTI test set #13

Closed binyang97 closed 4 weeks ago

binyang97 commented 4 weeks ago

Hi,

thank you for the work and publication. I have a question regarding to the results you show in the paper. Do you use any trick like test-time augmentation or model ensembling to further boost the performance?

Best, Bin

fengluodb commented 4 weeks ago

I don't use any tricks. I simply select the best result from multiple training runs.