Skip to content
/ STCPA Public

Official code for "Traffic Speed Imputation with Spatio-Temporal Attentions and Cycle-Perceptual Training" (CIKM'22).

Notifications You must be signed in to change notification settings

Sam1224/STCPA

Repository files navigation

Traffic Speed Imputation with Spatio-Temporal Attentions and Cycle-Perceptual Training

Accepted to CIKM'22.

Requirements

  • PyTorch 1.10.0
  • PyTorch Geometric 2.0.2

Usage

  • Download datasets (note that Chengdu dataset is from Didi GAIA, which is not public) from Google Drive.
  • Download pretrained models from Google Drive.
> cd STCPA
> conda env create -f stcpa_env.yaml
> conda activate stcpa
> pip install torch-scatter torch-sparse torch-cluster torch-geometric -f https://data.pyg.org/whl/torch-1.10.0+cu102.html
# Test on New York datasets.
> python test_stcpa_nyc.py

About

Official code for "Traffic Speed Imputation with Spatio-Temporal Attentions and Cycle-Perceptual Training" (CIKM'22).

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages