I could NOT get a guy in a prison gray pajama outfit like I see in those opening scenes. But, the general idea was fun getting at least something almost similar...
View attachment 3715159 View attachment 3715160
3.51 MB file on MEGA
mega.nz
I have no words Casshern. Absolute success. Im looking foward to see what is next. Congrats, mate.I could NOT get a guy in a prison gray pajama outfit like I see in those opening scenes. But, the general idea was fun getting at least something almost similar...
View attachment 3715159 View attachment 3715160
3.51 MB file on MEGA
Incredible! Truly inspiring.Brought these cuties to life using Wan 2.2.
View attachment 3722387View attachment 3722388View attachment 3722389
Images from this post
https://www.akiba-online.com/threads/photorealistic-ai-generated-images.2123060/page-47#post-4928279
Video
https://mega.nz/file/bJci2SCY#DA4s1PCWfZJW2bkH9KaF8R82cpPrxk1G9czEZFG8-1w
Hey, there, sorry for the late reply, was under the weather for a bit. All AI using Forge and an initial model for the text-to-image named cyberrealisticPony_v125 then resized/upscaled using a model named lucentxlPonyByKlaabu_b20 for image-to-image
Hey, there, sorry for the late reply, was under the weather for a bit. Good deal on trying ComfyUI, hope you're having fun with it like I am. It can seem like a lot, going strictly by the visuals of it, but once you use a workflow or two as-is getting the mentioned models in place then watching a YouTube video here and there, it starts to get easier to manage and change. Don't be afraid to move things around for things to make sense for you, it's 100% flexible as you know, so all the workflows out there are just laid out however the author did it. I tend to move things around to make them try to fit on one screen.
 | I've attached the workflow I used. Again, I moved things around, maybe in a non-standard way, but only because I know now how things connect and in what order by all the videos and articles and looking at other workflows. This one is using the Wan 2.2 high and low noise models. I got spoiled with LTX early on, but they've languished and haven't produced things as fast as the Wan group and community for it. They finally have something new about to come out, but by now I've moved on. Yes, it is faster, but the quality and prompt adherence is way better with Wan (maybe just for now). Here are the models/"lightning" loras I used. These are the GGUF versions, though, since the real models are over 20GB each and take a while to load for starters. If you're system can handle it you can try those, but you'll need to replace the GGUF Loaders with Checkpoint Loaders: Wan2.2-I2V-A14B-HighNoise-Q4_K_M.gguf Wan2.2-I2V-A14B-LowNoise-Q4_K_M.gguf Wam2_2_Lightning_high_noise_model.safetensors Wam2_2_Lightning_low_noise_model.safetensors For the prompt, I keep it simple as I've read more than one person say for image-to-video you don't have to give so much detail on what is already on the image, it knows there is a person/woman there, no need to describe her in detail or the surroundings for that matter. That's more essential for text-to-video. |
Here is a video based on the images you can find in the thread below.
Wow! What a time to be aliveMoved on to Christmas now that Halloween is over.
Had fun with Halloween images, now we'll skip Thanksgiving (here in the States) and head right to Christmas!
Nice model to use, and it's fast. GonzaLomo XL/Flux/Pony. I used the version 4 for this (gonzalomoXLFluxPony_v40UnityXLDMD.safetensors). It's up to version 6, I'm testing with that. Using the same seeds for these didn't have much of a difference, though. I created some videos of these using ComfyUI and the Wan 2.2 model. You can find...
Thank you for your reply.Hey, pal, no such thing as off topic on any threads I've started!
1. From a technical perspective, is it possible to extend or modify existing movie scenes using AI? - Yes, and that seems to get better as things progress. Originally the best way to was to extract the last frame of a generated video and use it as the source of the next video. But, that lead to image/video quality degradation and of course the AI had no idea what was going on before that last frame, so there would invariably be a jolt or sudden change of action regardless of the prompt being used. These days they have workflows that can use the last x number of frames from a video to give the AI an idea of the motion to guide it into producing a more seamless continuation, or as close as possible. And they now have some ways to combat the degradation somewhat.
2. I’ve noticed that AI videos can be generated either through APIs/web apps or locally on a personal machine.
Is there potential to earn money in this field, since hosted web-based solutions might be more affordable for users? Money to be made for hosting a site to produce videos, you mean? And charge for membership use? I'd say yes, but, I could only image (literally, because I don;t know and have never looked into it) there would be a huge(ish) upfront cost? You'd need to host the site and then rent or outright purchase a ton of GPUs so that your members don't complain about things taking forever to generate. And you would have to really lock things down so that no one can create illegal content in any way. That is a big liability.
3. Aside from bypassing restrictions or censorship, what are the main advantages of generating videos locally? Buying an appropriate GPU for your system would be at least one upfront cost to generating things locally. You could go big or step things up. I started with a much slower GPU and was happy as a clam until I found a deal on a 3060 and I've been happier still. That will pay for itself before long. Some out there might start with one site they pay for, then move to another or keep multiple accounts. That was me for a while when I was first interested in this. The biggest advantage is being able to produce as much as you want without running out of credits/tokens or whatever a membership on a site would cap you at before asking for more money or higher level memberships.
Bottom line, there's a magnitude more freedom doing things locally. And you're not locked down with things you can try or use. You are able to test and taste all kinds of models out there.
There can be opportunities, but eventually down the road that bubble may pop due to the exact thing we're discussing. Things will get cheaper, things will get easier, and most will start doing things locally, no need to pay for someone else to do things. Just an opinion, though, there's always a market for something out there!
