Here are some examples for the dust and scratch model that we are working on. If you have files to test, please upload them to the dropbox John shared above and send an email to lingyu.kong@topazlabs.com so I know to process and return your images back.
Before (top) and after (bottom)
Before (top) and after (bottom)
Thank you John and Lungyu for working on this exciting new feature. This definitely seems like a natural addition to Topaz and it eliminates the need to unnecessarily apply permanent or inferior algorithms to scanned pictures and film at the time they are scanned which forever tarnishes the archival value of the scan.
I look forward to this new feature which will likely compel me to add Photo AI to my current suite of Topaz tools.
Thanks again and keep up the good work!
Brian
An AI removal of dust and scratches would be great! Just be careful and leave an option that there are no other changes done to the pictures.
Hello everyone! If you’re interested in Dust and Scratch removal, I’d love to chat. Please send me an email at dakota.wixom@topazlabs.com with the subject line DUST & SCRATCH. Send me some example images you want recovered.
Shhhh. No spoilers! 
I would just like to add NEVER remove detected dust from the eyes of people, so many dust removal tools will remove the glint and reflections of people’s eyes, it makes them seem soulless and wrecks them.
Many scanners got an IR channel for detecting dust and scratches. In some programs photos can be exported with that data (like Vuescan). Could it be an idea to have the option to use that data too?
Any update on this feature’s release? Anyway to signup for early access? I need this very badly.
This enhancement is currently in Beta testing and will be available very soon in a future release for Photo AI that we are very excited to share with our users!
We’re happy to announce that Photo AI is released with a new dust and scratch tool
We will continue to improve the model and if it’s not working well for your photos we would be very interested to see examples. Please test it and write to us at support@topazlabs.com if you have feedback.
