Keras ImageDataGenerator brightness range

Question

In the docs for Keras ImageDataGenerator, there is reference to an arg called brightness_range (default None). The documentation says that this arg accepts

Tuple or list of two floats. Range for picking a brightness shift value from.

That is extremely ambiguous. Does it mean a float within range (0,255)? Or does it mean a float within range (0,1)? The lack of specificity suggests that both would work, but I don't know how this tool is meant to be used. Is there any further documentation out there or general best practices for using this arg?

Answer 1

As this article shows, 0 means no brightness (completely black), 1 means maximum brightness (no change), and anything above just makes the image brighter.