TeCSAR-UNCC / CARPe_Posterum

Code for "CARPe Posterum: A Convolutional Approach for Real-time Pedestrian Path Prediction"

Home Page:https://arxiv.org/abs/2005.12469

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

CARPe Posterum

A Convolutional Approach for Real-time Pedestrian Path Prediction
Presented at AAAI 2021

Required Libraries

PyTorch
PyTorch Geometric
See requirements.txt

Commands

bash download_data.sh #Will download and unzip dataset in correct location
python3 main.py

Citation

@misc{mendieta2021carpe,
      title={CARPe Posterum: A Convolutional Approach for Real-time Pedestrian Path Prediction}, 
      author={Matías Mendieta and Hamed Tabkhi},
      year={2021},
      eprint={2005.12469},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}

About

Code for "CARPe Posterum: A Convolutional Approach for Real-time Pedestrian Path Prediction"

https://arxiv.org/abs/2005.12469

License:Apache License 2.0


Languages

Language:Python 97.5%Language:Shell 2.5%