Updated the code to work on TensorFlow 2.0x and Python 3.

[thinktanktom]

Also added a resize program to resize all the images in the test directory to the same dimension to avoid possible errors.

[tyron]

@CodeSaint98 I'm trying to run your code, but I'm unsure on what's next after main.py finished execution. Can you clarify how to rebuild all the images after you found

automatic-watermark-detection/main.py

Lines 42 to 44 in fa14d6c

	# now we have the values of alpha, Wm, J
	# Solve for all images
	Wk, Ik, W, alpha1 = solve_images(Jt, W_m, alpha, W)

Any help would be greatly appreciated!

[thinktanktom]

@CodeSaint98 I'm trying to run your code, but I'm unsure on what's next after main.py finished execution. Can you clarify how to rebuild all the images after you found

automatic-watermark-detection/main.py

Lines 42 to 44 in fa14d6c

	# now we have the values of alpha, Wm, J
	# Solve for all images
	Wk, Ik, W, alpha1 = solve_images(Jt, W_m, alpha, W)

Any help would be greatly appreciated!

Hi @tyron . It's been a while since I worked on this so I don't remember too much. I remember running the code on jupyter notebook first because it was easier to see what was happening. I ditched this project because the accuracy was nowhere near good enough.

[kking124]

@CodeSaint98 out of curiosity, what project did you end up going with?

[thinktanktom]

I ditched it in the end because its a never ending cycle. As soon as we come up with something to remove watermarks, someone else comes up with something to counter it and vice versa.
I believe the methodology behind this implementation actually comes from a research paper designed to build more robust watermarks :P.
I happened to be playing around with this particular project as it seemed the most promising at the time. Resizing the images is just a small modification to help people run their images in bulk.