>>133539 Oh wow, what website you used for this?

I want to do a video like the screenshot an I have asked someone but no luck

now pixverse has sequencial video generation. And for what i saw its pretty good, we just need a way to manage to make an image of a deflated woman and a inflated woman

>>133545 Some app in an Instagram ad. Kling ai is really good. Start frame and end frame works well and I barely did anything to it. It would be great if someone did comic sequences and animated it on there

>>133801 I think Leonardo Ai allows the user to make an image and then edit it, its just changing the prompt and using the same seed btw. But its pretty complicated to still make the inflated and deflated woman look the same and the scene consistent, but when you do, the vids look epic. Btw Kling Ai its good but the generation is too slow, pixverse makes videos faster

>>133801 Same guy I tried to make this

>>133966 Looks very good, what ai you used?

>>133966 GREAT FUCK.

>>119848 >>133966 Only six months between these.

>>133991 We are entering a golden age! I can say the same about AI chat models. Half a year ago they broken the immersion with stupid shit responses very often and I had to regenerate 5-10 times to get somethingdecent. Today I am regenerating responses, because each of them is fucking great and I am having a hard time deciding which direction for the chat to go!

>>133966 Same guy here Did this last night

>>134059 wow, what ai did you used? Pixverse?

>>134061 what? i dont get it

okay, so, who's willing and able to train a hunyuan lora?

So, I've been playing around with pixverse AI, using it to give some videos a bit of extra spice, with varying degrees of success. However, I believe I have just peaked. Behold, the Female Creosote scene with an extra twist at the end.

Not sure where else to post this, but here goes. Jinglevellrock and several others came out on Twitter blasting AI artwork, and specifically targeted works using Spherify... which has led to Spherify's creator Floatful shutting it all down. So if you're wondering why you can't access the website anymore... here's your reason.

>>135218 I should also be more specific. Floatful came out and said they received threats in their DMs and their Disc ord, which came out because of these Tweets.

>>134059 >>133966 I rly want to do stuff like this can you tell the program u use?

"in a neon party with a letter that says Puffer Lounge, a beautiful black female bodybuilder with pixie haircut with deep eyeliner and deep black lipstick wearing a black bikini and tattoes sitting in a couch smoking a square device as her entire body gets fatter and becomes extremely inflated and obese wearing a black bikini and her arms becomes extremely inflated and obese and her face cheeks becomes extremely inflated and obese, with glossy skin" What if in a cyberpunk society there was a kind of Vape-Like Device called "Puffers" that causes the user to inflate like huge balloons? In here was recorded a video with a phone in a Puffer Lounge where the famous Wrestler Odessa used the eccentric Puffer for the first time

>>135253 btw if is possible someone can add sound please? or see if there's an Ai that adds sounds

but yeah, with enought promping we can make fake inflation scenes and stuffs like that, or something that might seem like a recorder video like what i did with strongfat inflation

>>135253 Is that the literal prompt you put into the app?

hey, it doesnt look bad. "in a neon concert stage, a muscular bulky female bodybuilder with short white hair with deep eyeliner and deep black lipstick wearing a black bikini and high heels, singing and dancing as she gets fatter and her entire body gets extremely inflated and grows and becomes a gigantic and extremely inflated extremely obese inflatable obese woman wearing a black bikini with extremely inflated obese inflatable arms with extremely inflated obese inflatable neck with extremely inflated obese inflatable face cheeks, latex skin" someone knows an ai that can add sound?

hey, Suno AI and Vozo works, i also used Krea. What do u think about this?

Managed to have a bit of success with vidu's image to video feature, but not much else

>>136014 Specifically, using the image to image feature

>>136014 What were your prompts?

hey btw, despite this might sound paradox or ironic, is there a way to make the inflation on vids not get naked? I tried in Vidu and Pixverse and often in Pixverse their bellies get bare without any kind of clothe ripping, and in Vidu they just straight up end up naked, wich makes the vids pretty odd. Someone knows how to make it the inflatee keep her clothes? Also, contribution in help: Pixverse is great in movement quality and manages inflation, but mostly focuses on bellies Vidu makes huge extremely inflated womens, but the movement quality is very bad. So, make a video in Vidu that you like, make it HD if you want tho, and then copy the first and last frames, or any deflated and inflated frames you like, and paste them in Pixverse for the Keyframe video generation, even you can make inflation slower if you pick as the last frame in Pixverse the frame of the middle in the last vid, and then make another vid with that last frame as the first frame and the last frame another. You can make it very long using this

>>136014 Good shit anon

>>136053 Here's have a few more of each.

Anyway, this all lead to me stringing a bunch of these together to bring a morph i did last year to life.

>>136108 NICE

>>135218 can you post the vid from the tweet in your screencap?

>>136108 I'm speechless. This is absolutely fantastic

>>136451 hey thanks haha. Here, have a lightly updated and slightly longer version!

>>136491 Btw how did you make it? I mean, what prompt u used? Also Pixverse is good for make inflation vids tho

>>136108 >>136491 AMAZING do you take requests or commissions?!

>>136493 This is several clips with different prompts all from Vidu that i stitched together with additional audio in a video editing so I don't think I can really help you with prompts haha. >>136506 Not really, sorry. This was just a little labor of love that took days to put together after getting inspired from some decent vidu prompts cherrypicked from dozens of failures. If i ever do something like this again it won't be for a while haha.

>>136510 well, maybe one single prompt, cuz every vid i try to do in Vidu, their clothes dissapear and turn into blobs..., so what you did for maintain the clothes still?

>>136512 Oh i ran into that constantly haha. First of all, i had a premade morph with her pants on as the end point of the video. Beyond that, I just kept emphasizing the clothes in the prompt, especially bringing up the "black pants" whenever i could. Even then, probably a good 75% of what i got still ended up with her butt turning pink or red before the pants snapped back on to match the end photo. Only reason i got what I did was because i just kept running the same prompt until it eventually gave me what i want. I'm not joking when i say what i ended up using was cherrypicked from dozens of vids that turned out bad.

Btw someone knows any good prompt for Pixverse? I feel like mines are too long, and dont even work as intended

https://www.deviantart.com/carlitosmurio

Yes

:3 Ai stuffer is fire

>>137190 Indeed, but they look too flabby for the forum...this fits better for BbwchannAi but good find. Btw it can make sequences or nah? Cuz if it can, we can make very great vids using pixverse or other vid generators

>>137213 Stufferai is based on Stable Diffusion, with a bunch of models you can choose from. SD can do a single pic with before/after, or maybe a set of comic-style panels -- if you're patient. But it has no "memory" from one generation to the next to be able to do a sequence. You can use the same seed number and modify the prompt to sort-of make a sequence, but keeping everything you want the same to stay consistent is challenging. Since you can run SD locally if your GPU is up to the task, it's a little easier and faster to do it that way then to use a shared service like stufferai. On the other hand, if you have the time to let SD kick out batches of images, some models like BiggerGirls V2 tend to put out a lot of sameface. Combine that with consistent prompting, and you can cherry-pick images into a sequence. For an example, I've put his one up on DA. It was assembled from several images that popped up after making batches here and there. There are captions to go along with each pic... A girl goes to an interview for an internship at a company that makes body inflation products, and she goes through stages of interviewing and getting pumped up to capacity. (I have some better ones on tap, but I need to finish up the captioning text and then I'll post them to my DA later, so there's more coming.)

>>137257 Another couple of images that came out independently, but I stitched together in a sequence with captioning... A performance artist does an installation that lets visitors inflate her boobs, belly, or butt according to their ideas of "beauty" -- which leads to her being massively inflated. So much that they have to move her out into the park/sculpture garden outside. (The face, hair and expression came out nearly identical -- and they were generated weeks apart.)