By Peiyun Hu, Aaron Huang, John Dolan, David Held, and Deva Ramanan
You can find our paper on CVF Open Access. If you find our work useful, please consider citing:
@inproceedings{hu2021safe,
title={Safe Local Motion Planning with Self-Supervised Freespace Forecasting},
author={Hu, Peiyun and Huang, Aaron and Dolan, John and Held, David and Ramanan, Deva},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
pages={12732--12741},
year={2021}
}
/data/nuscenes
.)conda
if possible), including torch
, torchvision
, tensorboard
, cudatoolkit-11.1
, pcl>=1.9
, pybind11
, eigen3
, cmake>=3.10
, scikit-image
, nuscenes-devkit
. (Tip: verify location of python binary with which python.)lib/grndseg
using CMake.preprocess.py
to generate ground segmentations precast.py
to generate future visible freespace mapsrasterize.py
to generate BEV object occupancy maps and object "shadow" maps. Refer to train.py
.
Refer to test.py
.
Thanks @tarashakhurana for help with README.