C4W2 Transfer learning

Hello there! In the assignment, I noticed that Image shape was from by adding (3,) to the image size (in this case was (160,160)). I understand that the input shape for mobilenet has to be 3D, but will adding a (3,) alter the image in any way? Thank you

The size of the images is 160 pixels by 160 pixels but they are RGB images, right? So you have a third dimension for the 3 color values per pixel. Generically they call that the “channels” dimension, since once you get beyond the input layer the values no longer represent colors. That logic is not modifying the images themselves, it’s just converting the shape into the appropriate description of the inputs.