Skip to content

Latest commit

 

History

History
94 lines (72 loc) · 3.1 KB

README.md

File metadata and controls

94 lines (72 loc) · 3.1 KB
Raw

VSeq2PC

The official implementation of A Novel Place Recognition Network using Visual Sequences and LiDAR Point Clouds for Autonomous Vehicles.

This work has been published on IEEE ITSC 2023: Huaiyuan Xu; Huaping Liu; Shiyu Meng; Yuxiang Sun

Overview

The figure shows the motivation of our cross-modal place recognition method. Given a query place in the form of an image sequence captured by a monocular camera, our method is designed to find the corresponding place from a LiDAR point-cloud database.

Pipeline

Get Started

Installation and Data Preparation

step 1. Please prepare environment as that in Docker.

$ cd docker
$ docker build -t your_docker_image .
$ docker run -it --shm-size 8G -p 1001:6006 -p 1000:22 --ipc host --name your_docker_container --gpus all -v ~/your_project_direction:/workspace your_docker_image  /bin/bash

step 2. Prepare VSeq2PC repo by

$ git clone https://github.com/lab-sun/VSeq2PC.git
$ cd VSeq2PC

step 3. Download data and checkpoints. Then arrange the folder as:

VSeq2PC/
├── checkpoints
    ├── model
    └── model_mae
├── data
    ├── gt_pairs 
    ├── img_desc
    ├── lidar_desc
    ├── pairs_folder
    └── pose_kitti 
├── docker
└── images
    .
    .
    .

Train models

In train.sh, please modify CUDA_VISIBLE_DEVICES=X, where X points to a specifc graphic card. Then, run the following command:

$ sh train.sh

Evaluate models

In eval.sh, please modify 'CUDA_VISIBLE_DEVICES'.

# Evaluate place recognition. We take F1 score as the metric.
$ sh eval.sh
# Evaluate loop closure. We take recall@K as the metric.
$ sh eval_recall.sh

Acknowledgement

We thank the fantastic works, RINet and Pytorch-Image-Classification, for their pioneer code release, which provides codebase for VSeq2PC.

Citation

If you use VSeq2PC in an academic work, please consider to cite our paper:

@inproceedings{xu2023novel,
title={A Novel Place Recognition Network using Visual Sequences and LiDAR Point Clouds for Autonomous Vehicles},
author={Xu, Huaiyuan and Liu, Huaping and Meng, Shiyu and Sun, Yuxiang},
booktitle={2023 IEEE 26th International Conference on Intelligent Transportation Systems (ITSC)},
pages={2862--2867},
year={2023},
organization={IEEE}}

Contact

[email protected]

Website: https://yuxiangsun.github.io/