In the Week 1 assignment the inception model input layer expects 299x299 image size and original image is of a different size. I don’t think the original image is properly resized in the original code. Does it mean we have to to resize ourselves the input image and resize the generated image back to the original size?
This assignment is more about feature extraction from the input image and as far as resizing the image, if you check down the codes, it is already done by developers as one needs to use argument recall variable name for image instead of input image.
Regards
DP