This repository is used to store the code of our paper "A Vehicle Trajectory Prediction Method Based on Multiple Graph Neural Networks and Transformers".
We use the NGSIM dataset and the HighD dataset for experiments. The data processing method is the same as that of CS-LSTM and STDAN. The original datasets can be obtained from the corresponding official websites.
Our model is saved in "stu.py"
You can try it out by running "main.py" < def train / eval >
Our configuration file is "defaults.py", and we can set parameters ourselves according to different data sets.
