Color coding

[janzem1]

First of all: Thank you for your infographic. I think it's amazing.

Can you tell me what's the difference between the red (like AlexNet) and green (like RCNN) boxes?

Thanks

[Nikasa1889]

Hi @janzem1,

First, CNN object recognition approaches are orthogonal to the CNN architecture. You can use any CNN architecture for any CNN object recognition approach.

Second, the development of object recognition and CNN architecture for image classification have been going in parallel. So in the "MiniMap", I use red box to mark a milestone in CNN architecture development, and green box for a milestone in object recognition. These boxes are also used in the References section, linking the milestone with the paper introduces it. There is also blue box for nice informational references.