Has potential but it’s not there yet. It has 3 modules so far – a placeholder indicates that additional modules are coming – and those three didn’t impress. Using one of my usual test images, a scanned 35mm 400 ISO photo print, the overall enhancement it offers make the photo, especially its film grain, look noticeably worse. The enlargement module wouldn’t work, with a message that the PC has insufficient memory – I *think* it might really mean graphics memory [I was running it in a VM], and it *might* also be related to a driver issue with the app itself. The background removing module basically followed the mask I painted on exactly, without evidence of anything AI… like many of these tools you can paint the areas you want to keep & those to get rid of. Ideally the software uses that as a hint and finds the object edges, accurately removes the object(s) and in many cases fills in the empty space, or in this case removes the background. Photoins just seemed to just follow the paint I applied exactly.
When you visit the site in the readme.txt file there’s 3 text boxes to fill out, first name, last name [I used an initial], and email address. That gave me a second page with the activation key & a download link. The small setup app I downloaded was a downloader, with the real setup file, at just under 1GB downloaded to Windows Download folder. That setup file was automatically run, adding the program’s folder, at around 2.4 GB, with new folders in ProgramData, Users\ [UserName]\ AppData\ Local & Roaming. The registry mainly got two keys, one in HKCU & one in HKLM. Note, I could not fully monitor the installation since this app is 64-bit only. In my Win7 VM it showed an error that only signed drivers could be installed. In my Win10 VM there was no such message, but the driver did Not install either. It’s *possible* that extracting the full setup file using Universal Extractor *might* work to get an .inf file that might allow installing the driver, but I did not attempt that. Photoins also gave a 2nd warning message that a file was missing so it could not run in my Win7 VM – it **might** be related to a hack I applied to allow Win7 to run on newer, unsupported hardware.
Using AI for all things photo related is a big thing nowadays, from deep fakes to editing/enhancement to Google’s new project creating images based on descriptions in plain text, e.g., a dog dressed as a firefighter. The hard part is training the AI, which requires HUGE amounts of data. Leawo’s Photoins will probably get there someday, while some editing apps offer similar, AI features, and DVDLab’s Photo Enhancer AI [a previous GOTD] is already there & continuing to improve.