NEW 1 minute long animated Video

[Kube²]

I uploaded my cherry 2000 AI enhanced stills, there is also a few production pics as this one











here is a little preview of the batch












the rest of them are here :

https://www.fembotwiki.com/index.php?title=Kube%C2%B2

[dubhdanaidh]

Thank you Kube. Always love your work.

[D-wolfmaster_2022]

You're a really wizard on these re-enhancement jobs.
Long live Kube2!

[Kube²]

Hey all !

CLASSIC CANDY DETAILLED FACE RE-HENHANCED IN 4K FROM FECTION (Candy on the rocks video)

Today I continue my series of community classics reworked and boosted with AI with something a little different, after all THE Candy from our dear Fection is also a classic. We may or may not like the move to 3D (I really liked that evolution) but we cannot deny that this move gives a more physical and real aspect to our dear Candy. It's much easier for an AI tool to extrapolate and add details but without giving it the semblance of life she's not supposed to have...

4K are version is on the wiki


IMAGES REMOVED BY ADMIN

[dubhdanaidh]

Thank you Kube. I always love your work.

[fection]

Nice work!

[Kube²]

I plan to recreate my old manipulations using AI. Here are some examples.

Basically you ask the AI ​​to use a single image (in this case a manual image created by me years ago). Then you specify via a cleverly elaborate prompt the treatment to be given to this image.

Here is a simple example :

IMAGE REMOVED BY ADMIN

Here is an other one that allow variations but is still based solely on human work :

IMAGE REMOVED BY ADMIN

I think this approach should not be considered AI art, since the AI ​​is used as a sort of advanced filter applied to an old manip created manually and created by myself, at the time it met all the requirements to be published on the wiki that should still be the case

If it turn out the moderation allow me to do this kind of experiments I also promise to avoid image flooding and I will follow a simple rule, a different manual manipulation will be at the origin of each image post-processed with the AI I post here and a link to my original image can be provided (if I published it here I give the old link and if it is a manipulation never published (I have many) the no AI version can be provided to the moderator if they request it)

I think I am the first here to have an approach of this type, I would like moderation to give me their opinion on this approach which I consider ethical and in line with the rules of the wiki.

Basically, this is not AI arts, It's human made manually edited picture with in AI filter apply to it then re-edited/enhanced/corrected manually again to fix weird AI filter shit and junk, then posted here.

What do you think ?

[sqr149]

This seems like a good approach to me. The original work is yours and you direct the actions to improve it so it is pretty much acting as a very good Photoshop tool to improve the image quality and fine details. Since it is almost all manual it avoids the flood too. As long as the original, pre filtered, is uploaded alongside the final, that would help to avoid a potential flood.

The only point I will say could be used against it, which I personally don't think is a problem, is that the improvements that to filter does are not 100% yours and were trained to do that off of other works and just copies your style in the current work. However, that is how many Instagram filters work to add effects or features to it so it is a little bit of a gray area. What art doesn't take inspiration or techniques from others at some point?

[Kube²]

I continue to explore the "ethical" creation process, but i'm going for something more extreme.

Now instead of staying very close to do original pic as I show in my last post here, I totally transform it into something totally new.

It require several step, for this manip there is 5 steps :

-Approximate and rough manip of the original pic so the AI understand what you want
-AI generation on the maniped pic with a prompt
-Creation of the new setting with a prompt
-Approximate and rough integration of the girl in the new setting (the AI will perfect the integration in the following step)
-Final integration and rendering with a new AI pass and a prompt

Rare thing to note I share for you some of my favorite prompt here. feel free to use, but be aware those prompt worth nothing without the image editing work before

IMAGE REMOVED BY ADMIN

[NatalieBayer]

Hey, I just removed the images from this post. I know you were simply showcasing some images, but we do not allow any AI generated images on the wiki/forum. I understand that you feel that since you made the original image and then put it into an AI generator that it becomes ethical work, but that still interacts with an AI image generator that has unethically scraped images from somewhere.

[Kube²]

Hey, I just removed the images from this post. I know you were simply showcasing some images, but we do not allow any AI generated images on the wiki/forum. I understand that you feel that since you made the original image and then put it into an AI generator that it becomes ethical work, but that still interacts with an AI image generator that has unethically scraped images from somewhere.

I was expecting this (even if I was secretly hoping it could be somehow okayyyish...).
It's not retaliation but I will not post any new manipulation here. Nothing against people here and policy of this place but after spending time into understanding how AI can assist me in my creation process i'm simply not interested in going back into classical image manipulation, that would feel like going backward for me and I don't like this feeling. I simply don't like to limit myself when I know I can do more.

By the way I will continue to post AI improved still from TV and films as it seems that it's okay.

[NatalieBayer]

I was actually unaware that any screenshots you were posting were also AI enhanced. I would say that for the same reasons as your manip process, that those should also not be posted here.

[Kube²]

Could you wait before eventually removing the TV and series stills because I talk a lot with Spaz at the time about what I was doing and he tell me it was okay. I have coordinated with him and even updated some of the pages he has created himself (I don't remember if I talk with dollspace and you about it but in my mind Spaz is the one to talk to when its about posting stills, he literally produced 99% of all stills actually on the wiki)

Also take into consideration that as far as I know there is not really any other place on the web to upload AI enhanced stills from fembot TV and film classics.

[NatalieBayer]

As I am sure you saw in DMs the enhanced screenshots are fine, I haven't removed any of them.

I'm not sure what Dollspace has to do with this discussion though, could you clarify that?

[Kube²]

don't know... Dollspace is the only one of you that is tag as moderator and this is about moderation, so... but I don't know how you work internaly

edit : i saw spaz message. its a good thing

[Kube²]

good news everyone !

Now that it made been clear that using AI to digitally enhance a lower quality video screencap to make something HD when it previously wasn't is legit in this place i will continue with it, starting with one of the finest Fembot TV show ever done in my opinion but also one of the hardest to find in acceptable quality.

Personaly I've never find this old "boxed in" episode from TV show perversion of science in more than 360p.



enjoy !







And there is more !



I've got a brand new deviant art page for all the new photo manipulation and AI recreation of my old work


it's here :
https://www.deviantart.com/kube-square

Most of my new image manipulation include AI at one step or an other and because I won't give up the tech now I'm starting to understand it better I need a new solution to showcase my work. Unfortunatly it can't be here anymore (sad). So this is the logical solution : my new edit and image manipulation are going to be on Deviant and Fembot AI enhancement TV/films screencap are going to stay here. This is the way.

Go watch that DA page, there s already a group of new manips and there's going to be a lot more really soon.








And now a small selection of the last batch from boxed in

[Kube²]

Here is an AI video of my personal Valerie 23 in maintenance

https://www.fembotwiki.com/index.php?ti ... _video.mp4

she likes maintenance. hope you like it too !

for moderators :
The video is AI but it is generated used as it only input a 100% manual manipulation (i posted it on the wiki) its easy to see its the only input for the video.
I thinks its of with the rule here but worst case contact me by pm before deleting

[Kube²]

Here is a 1 minute long animated Video, spend a lot of time on this one, enjoy !

LORE : Testing basic functions

This obsolete robot girlfriend had been ruined for a long time but her boyfriend/owner couldn't bring himself to scrap her. After many years of trying to bring her back online, he finally found a way to mimic her old artificial personality on modern hardware. He turned it back on for the first time in years. The robot immediately began testing its chassis and took its first steps in years while her boyfriend explained to her that after spending so many years disabled, she needed maintenance and repairs so much that he had to remove most of her artificial skin to work on her more easily.

https://www.fembotwiki.com/index.php?ti ... d_unit.mp4

streamable link for those who prefer it
https://streamable.com/j6pbve

(for moderation it is animated with AI but on the basis of an none AI photo manipulation I've done before, and it's the only input used, thanks to contact me by MP before deletion so i can have the possibility to explain myself, those kind of stuff is hard to do and I share it for free, because i'm a cool dude... so be cool with me too pliiiiiiiiize)

here is the NOT AI input (old manip from 2017 I've re editied (manually) to make it more clear and easy to "understand" by the IA (mainly improving contrast and shading so the AI tool understand better what is inside and what is outside and the multiple layers of the image)