search

ZikangZhou / HiVT

[CVPR 2022] HiVT: Hierarchical Vector Transformer for Multi-Agent Motion Prediction
https://openaccess.thecvf.com/content/CVPR2022/papers/Zhou_HiVT_Hierarchical_Vector_Transformer_for_Multi-Agent_Motion_Prediction_CVPR_2022_paper.pdf
Apache License 2.0
577 stars 115 forks source link

What do "semantic attributes" represent? #32

Open Cram3r95 opened 1 year ago

Cram3r95 commented 1 year ago

Hi,

Your work looks quite interesting. I have read your paper and I am wondering what do "semantic attributes" represent here:

imagen

And here:

imagen

Best,

Carlos

SupOne11 commented 11 months ago

I guess this is Lane semantic information, including whether it is a right-left lane, whether it is an intersection, and whether it is controlled by traffic lights.