Input Data Requirements

[MarioCodeMatuschek]

First of all, thank you very much for open-sourcing the code to the YoloR framework.

However, when applying the system in practice one question came to my mind. Does the network favor any specific types of images as an input?

Further questions that appeared at the back of my head are:

• Is nature favored over other indoor-type images?
• Are specific type of objects easier detectable than others?
• Are there any general requirements that should be imposed on the input data such as ideal image size, ratio to object size, etc.?

I am looking forward to an insightful answer.

Best regards