This repository contains the official PyTorch implementation of our paper:
[IROS 2024] PP-TIL: Personalized Planning for Autonomous Driving with Instance-based Transfer Imitation Learning [arxiv]
-
[20/08/2024]: Our code is released!
-
[04/08/2024]: We release the PP-TIL paper on arxiv!
-
[30/06/2024]: PP-TIL is accepted by IROS 2024🎉!
Download the dataset from Waymo's official website: [web]
Select the dataset folder as follows:
waymo_open_dataset_motion_v_1_2_0/uncompressed/scenario/training_20s
We employ serial numbers ranging from 0 to 799 as the training set and those from 800 to 999 as the test set.
If you have any questions or suggestions about this repo, please feel free to contact us (linfangze2023@email.szu.edu.cn).
If you find PP-TIL useful in your research or applications, please consider giving us a star 🌟 and citing it by the following BibTeX entry.
@article{lin2024pp,
title={PP-TIL: Personalized Planning for Autonomous Driving with Instance-based Transfer Imitation Learning},
author={Lin, Fangze and He, Ying and Yu, Fei},
journal={arXiv preprint arXiv:2407.18569},
year={2024}
}
PP-TIL is based on the following projects: DIPP, MultiPath++. Many thanks for their excellent contributions to the community.