search

TempleRAIL / drl_vo_nav

[T-RO 2023] DRL-VO: Learning to Navigate Through Crowded Dynamic Scenes Using Velocity Obstacles
https://doi.org/10.1109/TRO.2023.3257549
GNU General Public License v3.0
113 stars 7 forks source link

Is camera image used in simulation? #11

Closed caomuqing closed 8 months ago

caomuqing commented 8 months ago

Hi, this is not really an issue but just a question. From the simulation, I don't see any zed camera image or depth data being used, it seems that the pedestrian's information from the pedsim is fed into the network directly. So, the simulation does not include the image encoding part described in the paper?

zzuxzt commented 8 months ago

Sorry for the inconvenience, the current open-source code does not include the Yolo detector and MHT tracker, and the pedestrian information comes directly from the pedsim_simulator, which makes the simulator run faster. We did not open source the Yolo detector and MHT tracker code because the MHT tracker required some commercial license packages, which is a bit complicated. But when we clear these commercial restrictions and update the code, we will consider open sourcing the same preprocessing pipeline as in our paper.

caomuqing commented 8 months ago

I see, thank you for your information!