Leonardo AI still remains problematical in many respects at creating still images but it can still knock them out of the park every so often, as in this image. Lovely image, just about perfect as is, except for a suspicious blob of flesh behind the woman's left hand, and some wonky looking fingers on her left hand. Both very easy to fix, and what a beautiful image otherwise, except for that very straight border between the stream in the back and the foreground green and... oh, hell better stop talking. So I tried a motion video based on this image. And boy, howdy...
What a great video. My prompt was simple, "A beautiful woman lying on a white sofa smiles and waves." Came out great! Some minor morphing
I tried a motion video based on the Second Life avatar rather than the Leonardo AI interpretation of the avatar, and got some pretty good results, nice facial expressions and hand movements, but, well, you'd best just see for yourself.
The sofa on this one is doing about as much moving around as the woman lying on it, which is unusual behavior unless the woman is lying perfectly still, which this woman definitely is not. And what's with the green geometric grid behind the sofa? It looks like the AI program created a perspective grid for the grass and then said "Ah, fuck it!" and just let the grid stand in for the grass. I know just how that AI felt.
Still, despite the moving sofa and the grass grid, the woman's animation was excellent, much better than usual. The hands look a little sus now and then as they move but not enough to make the animation look bad, and the facial expressions are excellent, as well as the generally honest nudity of the body. But the absolute best part is the way the animation captured the gag and collar. Leonardo AI is very prone to misinterpret gags, because putting the term "gag" in a text prompt makes Leonardo AI refuse to do the animation and give you a warning about, and I kid you not, "sex and violence" in the image.
So mostly I don't mention collars and gags and cuffs and shackles and hope for the best: "A beautiful woman waves and smiles." And mostly what I get is wierd-looking gag-like things and instead of collars I get necklaces. Cuffs morph into bracelets, generally silly-looking ones. Ankle shackles become ankle bracelets and ringlets and sometimes shoes. (Which, granted, is often an improvement over whatever the hell Leonardo AI does to feet.)
But the collar and gag on this cutie are dead-on, pretty much the same as the ones on the SL Avatar except they're in motion. I'm sure there's a lesson in there somewhere for all of us, but I'm not at all sure what it is.
Let's close with an example of a Leonardo AI animation handling a gag badly:
Ladies and gentlemen, may we present the Duck-Billed Slavegirl! Woo-hoo!
But overall, for nekkid cutie animations, Leonardo AI Motion is still kicking Kling AI's ass big time. Because except for that duck bill excuse for a panel gag, that's one hell of a finc animation of a naked woman writhing enticingly on a sofa.
