Honest question, is there even a legitimate use for this specific tech?
It's not illegal to remove watermarks from photos. Ethics is another thing.
Watermark segmentation
11–20 of 22 posts
Re: Watermark segmentation
#12Honest question, is there even a legitimate use for this specific tech?
Re: Watermark segmentation
#13Re: Watermark segmentation
#14Re: Watermark segmentation
#15Earlier quoted context omitted.
It's not illegal to remove watermarks from photos. Ethics is another thing.
Incorrect. Removing the watermark constitutes a derivative work. To distribute this work you need permission from the copyright owner to be legal. This you will almost certainly not get since the point of watermarks is to keep people from stealing copyrighted material.
Re: Watermark segmentation
#16Honest question, is there even a legitimate use for this specific tech?
Re: Watermark segmentation
#17Honest question, is there even a legitimate use for this specific tech?
Re: Watermark segmentation
#18Earlier quoted context omitted.
Incorrect. Removing the watermark constitutes a derivative work. To distribute this work you need permission from the copyright owner to be legal. This you will almost certainly not get since the point of watermarks is to keep people from stealing copyrighted material.
When in the process of removing the watermark do I distribute the work?
Re: Watermark segmentation
#19Their first example at https://www.clear.photo/en is absolutely terrible. I assume a showcase would show "good" results, but they display a complete failure. - Incorrectly identifies areas for inpainting. You can see this with the figure, a lot of detail, not obscured by the watermark, is erased and then redrawn. This leads to a totally distorted look. The belt just disappears into nothing, the cloth just becomes a g…
What would be the state of the art?
Re: Watermark segmentation
#20This is technicaly impressive, but I wonder if this could be put to a use which is generally more constructive. Like maybe removing stains from scans or red eye from pictures.
Look at their first example on: https://www.clear.photo/en How is this technical impressive? It fails at segmentation and it fails at inpainting. I presume for a commercial product you would but a successful result front and center.
Now they replaced it with an image where the inpainting just needs to fill in a gradient. Which is of course trivial.
Why do you not make your product better instead? Obviously the first one was what the customer should expect from your product. Also look at the top left tree! The segmentation still fails to correctly identify the watermark.