two-stage-gan-in-trajectory-generation's Introduction

Large Scale GPS Trajectory Generation Using Map based on Two-stage-GAN

Python impletement of paper Large Scale GPS Trajectory Generation Using Map based on Two-stage-GAN (Link)

Introduction

we propose a map-based Two-Stage GAN method (TSG) to generate fine-grained and plausible large-scale trajectories. In the first stage, we first transfer GPS points data to discrete grid representation as the input for a modified deep convolutional generative adversarial network to learn the general pattern. In the second stage, inside each grid, we design an effective encoder-decoder network as the generator to extract road information from map image and then embed it into two parallel Long Short-Term Memory networks to generate GPS point sequence.

Result

We evaluate the synthetic trajectories in terms of their similarity to real data, i.e., distribution of overall GPS coordinate, distribution of trajectory sequences length, distribution of trajectory distance, top-N visited places and road networks matching accuracy. And we compare our result with the previous benchmark.

JS distance of distribution

Model	$p_o(r)$	$p_s(l)$	$p_d(l)$
FTS-IP	0.413	0.182	0.187
LSTM	0.633	0.058	0.140
TSG	0.100	0.139	0.136

Visualization of road network matching

Train

Dataset

Trajectory data in Porto, available on Kaggle

Prepare the data

Transform trajectory data into grids format pre_process/process_trajectory_data/to_grid.py
Prepare the corresponding map images:
- go to pre_process/map_generation/
- run screen_shot.py
- run cut.py
- run merge.py

First stage GAN

go to First_stage_gan/.
run:

python WGANGP.py \
--dataroot ./grid32/ \
--labelroot ./traj_all_0115.txt \
--outf ./output \
--batchSize 64 \
--n_critic 1 \
--netG ./output_IN/netG_epoch_320.pth \
--netD ./output_IN/netD_epoch_320.pth \
--cuda \
--start_iter 320 \
--niter 350

Second stage GAN

go to Second_stage_gan.
run python train.py.

Generate trajectory data

Coarse result generated from First stage GAN

cd First_stage_gan
python generate_fake_data.py --large_num 200 --model_path ./output_IN/netG_epoch_260.pth --output_path ../output_generated_coarse

Final result TSG/pred.py

Configurations
- step_1_output path to the result of first stage GAN
- map_dir path to the map data
- checkpoint model result of second stage GAN

Citation

@article{wang2021,
    author = {Xingrui Wang and Xinyu Liu and Ziteng Lu and Hanfang Yang},
    title = {Large Scale GPS Trajectory Generation Using Map Based on Two Stage GAN},
    journal = {Journal of Data Science},
    volume = {19},
    number = {1},
    year = {2021},
    pages = {126-141},
    doi = {10.6339/21-JDS1004},
    issn = {1680-743X},
    publisher = {School of Statistics, Renmin University of China}
}

two-stage-gan-in-trajectory-generation's People

Contributors

Stargazers

Watchers

two-stage-gan-in-trajectory-generation's Issues

question

Can you provide the data style of the data set after processing，for tj = np.load('./data/9.npy', allow_pickle=True)
pre_process\process_trajectory_data\to_grid.py the lat and lon variables are undefined and are min_ Lat and min_ Lon
for i in t1:
a = int((float(i[1])-lat)//d) # lat
b = int((float(i[0])-lon)//d) # lon
t2.append([a, b, i[2]])

File / File Preprocessing Missing

Dear authors,

When running pre_process/process_trajectory_data/to_grid.py is looking for the file ./data/9.npy and something similar happens with generate_traj_plot.py . Would it be possible to have a script/specs on how to transform the trajectories in the npy files and in which path we should save them?

Thanks in advance

xingruiwang / two-stage-gan-in-trajectory-generation Goto Github PK