Convolutional model isn't accepting the image array's input even though provided with right shape

Question

I am taking two patches of size (96,96,3) from an image and feeding it to a convolutional network . When I try to feed these patches as input , it is giving me this error.