Explanation of the dataset

[BonanWang]

I noticed that in Seq-TFD, there are categories like single-infrastructure single-vehicle and cooperative-vehicle-infrastructure. Among them, cooperative-vehicle-infrastructure further includes vehicle-trajectory and infrastructure-trajectory. I would like to ask, what are the differences between them?

Additionally, I have noticed the cooperative-trajectory directory, but it wasn't used during prediction. Why is that? What is its purpose?

Moreover, if cooperative-trajectory represents a cooperative view, why does the data preprocessing involve a fusion environment?