The training mode of "READ"

[booker-max]

You used two datasets, namely KITTI and Brno Urban datasets. The KITTI dataset contains three scenes: KITTI Residential, KITTI Road and KITTI City. The Brno Urban dataset contains three scenes: Left side view, Left front side view and right side view, I want to ask about your training mode on the two datasets?

1. Did you train a model in Kitti based on all the data in the three scenarios, or did you train a model on each of the three scenarios and ended up with three models, I'm confused.
2. In the same case with the Brno Urban dataset, did you train one model or three models?