search

hova88 / PointPillars_MultiHead_40FPS

A REAL-TIME 3D detection network [Pointpillars] compiled by CUDA/TensorRT/C++.
GNU General Public License v3.0
279 stars 88 forks source link

Nvidia Jetson devices support! #9

Closed YoushaaMurhij closed 3 years ago

YoushaaMurhij commented 3 years ago

Great work! have you tested this model on Jetson devices like TX2 or Xavier ? or Do you have any plan to implement it in the future ?

hova88 commented 3 years ago

Great work! have you tested this model on Jetson devices like TX2 or Xavier ? or Do you have any plan to implement it in the future ?

Yep, I tested it on Xavier. As long as the environment is consistent (TensorRT 7.1.0+，CUDA 10.2, onnx2trt ), running this repo is basically no problem. But you need to note that there are many imperfections that you need to improve.

likegogogo commented 3 years ago

@hova88 @YoushaaMurhij Hi.PointPillars_MultiHead_40FPS with trt model in xavier, how much inference time takes?