[AI Technique] Banter, shared ideas, and frustrations

[xs70]

And here I am again with my questions

Welcome back.

I understand this is an animation in Wan2.2?
Can this be achieved with a simple prompt or do I need to use specific LoRA?

WAN2.2 Image to Video (I2V). Just the prompt, no specific LoRA's.

[GraekrasBlight]

Can those of you who are knowledgable about Comfy UI and other local server AI generating apps post links to your DA galleries or other places online to check out your work?

I have been doing all of my work using various online apps, but I have just hit a wall. Here is my DA profile:
https://www.deviantart.com/graekrasblight/gallery

I just posted a video a few days ago that I'm happy with - hit a new level of realism. It's for watchers only, so you'll need to add me to your watchlist.

I used the aitubo image reference video generator for all of the shots. It was uncensored and unrestricted, and even though results were inconsistent I was able to get enough to put together the video. However, overnight they changed the restrictions and now even partial nudity immediately gets censored. I don't know if there is an alternative to aitubo - most are pretty bad or they are highly restictive (I've also used kling and hailuo - decent results, but very restitcive). Aitubo does have their A4 uncensored generator, but it is really bad.

To run locally, I will need a hardware upgrade and I'm sure there will be a learning curve. That's why I'd like to see other work that people have done locally using Comfy UI. I'm curious if I'll be able to figure out a workflow that can generate results similiar to what I was getting with the aitubo image reference generator.

[Kelderek]

Can those of you who are knowledgable about Comfy UI and other local server AI generating apps post links to your DA galleries or other places online to check out your work?

Since you're watching me on DA, you must have found me already... but here is the link to my DA gallery:

hwww.deviantart.com/mrkelderek

I use Krita AI Diffusion, which is an AI plugin for Krita, basically embedding Comfy UI into Krita:

https://github.com/Acly/krita-ai-diffusion

The AI model used is Flux.1 Dev, but since the standard Flux model is censored, I use "NSFW trained" checkpoints that allows nudity.
The ones I mostly use are GonzaLomo and STOIQO Afrodite:
https://civitai.com/models/1513492/gonz ... d=1883195
https://civitai.com/models/207101/stoiq ... nId=897489

[GraekrasBlight]

Can those of you who are knowledgable about Comfy UI and other local server AI generating apps post links to your DA galleries or other places online to check out your work?

Since you're watching me on DA, you must have found me already... but here is the link to my DA gallery:

hwww.deviantart.com/mrkelderek

I use Krita AI Diffusion, which is an AI plugin for Krita, basically embedding Comfy UI into Krita:

https://github.com/Acly/krita-ai-diffusion

The AI model used is Flux.1 Dev, but since the standard Flux model is censored, I use "NSFW trained" checkpoints that allows nudity.
The ones I mostly use are GonzaLomo and STOIQO Afrodite:
https://civitai.com/models/1513492/gonz ... d=1883195
https://civitai.com/models/207101/stoiq ... nId=897489

Your work is great. Thanks for the info. Do you do video? I really need to figure out video locally because the web based apps have all changed in the past week. It sucks that the model that enabled my last video is now just gone.