Romeno is a small town along the road. I wouldn’t have included this image, if it were not the first on my blog where I’ve removed something with Lightroom’s new generative AI.
Is it good? Well, it works. It produces acceptable results, arguably much better than the context-sensitive fill before, but it does not match the resolution of the image. It’s good for small patches, and you can help it along by painting some noise over the patch to fake more detail. Here I’ve removed two small but obnoxious blue and white signs marring the two church towers, and two more signs in between, part of a parking car on the right side and a car and a tractor at the end of the street.
I won’t do that regularly, but this was a good image to try generative AI. The result is really ok, no cats were hurt and apart from the signs, there are days you can see the street like this. Just not when I was there.
It’s a fine line. We’ll discuss the ethics and a possible framework soon.