ispc-lab / GSC

Spatial-Temporal Graph Learning with Self-supervised Spatial State Module for Traffic Accident Anticipation

GSC: A Graph and Spatio-temporal Continuity Based Framework for Accident Anticipation

The repository contains the source code and pred-trained models of our paper: GSC: A Graph and Spatio-temporal Continuity Based Framework for Accident Anticipation

Architecture

The overview of the network is showed below;

Prerequisites

• Python 3.6
• Pytorch 1.7.0
• Pytorch-Lightning 0.9.0
• Other required packages in requirements.txt

Getting Started

Create conda environment

conda create -n sspm python=3.6
source activate sspm

Install the required packages

pip install -r requirements.txt

Downloading MASKER_MD dataset and unzip it

• Access the datasets by BaiduYun[Passwards: qo81], and unzip it.

• Change the data_root of configs/config.py to your unzip path;

Train the Model

• Run the following command in Terminal:

  python run.py --train ./configs/config.py

Test the Model

• Change the test_checkpoint of configs/config.py to your model

• Run the following command in Terminal

  python run.py --test ./configs/config.py

Visualize