AI and Regenerative Imaging

The #AI processes around #RegenerativeImaging have come a long way in the past 6 months, to the extent that you can’t take most of the photos you see on the internet at face value anymore. As an exercise, I took the base version of #gemini – available FOC to any user out there – and fed in a photo of Karyn I took at Kagga Kamma Nature Reserve on one of my recent shoots.

That original image then went through 5x PROMPTS (which I’ve embedded in each of the images, if you’re interested) all executed in less than 5 seconds. None of this is groundbreaking, but maybe just see this post as a simple reminder that the visual landscape – and #photography itself – has changed forever.

Paid versions of #AI software – or those iterations more specifically focussed on #ArtificialImaging – allow enhanced control of all the aspects of any visual image. Or you could just spend more time, for example focusing on the blood-moon, adding texture and depth to make it look more realistic.

In @META and a host of other programmes, you can now also take any specific photograph and animate it. There are glitches, but most of those will be tweaked and perfected in another six months as the machine-learning speeds up. I’ll do a post of this #animate process at some stage …
Ending off with a Public Service Message: when you see a video of an eagle grabbing a baby while a jack-russel fights it off, or a cougar adopting a domestic kitten, please take a moment. And I don’t mean an #AhCute moment, where you re-post it and cause me to have an #AgFok moment.

 

Click here to view the pics

Search
Archives