A workflow that could change the clothes in a video by using VACE.
Free try out link: https://www.runninghub.ai/post/1963409959855820801
All the usage and model details are within the workflow notes.
The video generation is resource-heavy. If your local computer is unable to run this workflow or you just want to see how good the model is and check all the workflow parameter settings before downloading, please open the link above to run it online for free on RTX4090. Just click and run. No brain-racking local setup.
1000 credits upon signing in using the above link(One gen only takes about 10-30 credits). With an extra 100 credits on daily login.
Description
FAQ
Comments (16)
are we supposed to guess how to use it?
Please read the description first. All the details are within the workflow.
yeah. a screenshot of the workflow would be nice.
Edit: its just vace with clothing segmentation
非常好，希望添加對長視頻的處理(處理一段約略1-2分鐘的視頻，分成多段處理，合併)
谢谢， 分段处理估计效果可能不太好，可能会有一致性问题，我在考虑直接用上下文衔接，就是跑的时间会久点。
Civitai needs a filter for affiliate marketing posts ...
The only issue with these posts is lack of description and more example renders. But in fairness they provide potentially useful workflows, as in this case (given it works as well as the example). Flooding is a far bigger issue (currently we get floods of a dozen or more anime posts from the same entity all at once- what happens if this becomes a hundred or more?) Personally I hate how 'naughty schoolboys' can flood the example render section of a post with dozens of near identical 'sexual' images- would it be so hard for this site to limit non-authors to say 10 images- after all a chain image can contain a number of individual images while taking just a single spot!
Thanks for your feedback. The thing is I use an old RTX 20s graphics card. so I think it might help for the people from a not so rich area to use free service if they just want to take a look. and for the lack of examples. I made a lot . but civitai got policy restrictions on that. so I can only pick some. And since it is a workflow. Better to give the instructions beside the nodes,so you don't need to go back and forth to check how to use it?
A gguf version would be nice
thank, I will keep it in mind.
having this problem in runninghub
cusolver error: CUSOLVER_STATUS_INTERNAL_ERROR, when calling cusolverDnCreate(handle). If you keep seeing this error, you may use torch.backends.cuda.preferred_linalg_library() to try linear algebra operators with other supported backends. See https://pytorch.org/docs/stable/backends.html#torch.backends.cuda.preferred_linalg_library
this may due to the server issue. you could try again later
Great result. Thanks.
If you plan try run this workflow on local comfyui, with another installed custom nodes, i will reccomend make clean version comfyui. Really a lot dependecies conflicts.
Also i run that on 5060 16GB VRAM 32RAM and multiple errors with low VRAM. After i increase block swap - it helped, but taked really a lot of time for render.
cant get started becasue it cant find these:
RH_LLMAPI_NODE(In group node 'workflow>Prompt Refinement')
workflow>Prompt Refinement
Your workflow is Great, but there's one problem: when the clothes changes, the area that was previously a mask but is now skin doesn't have the same color as the skin that wasn't covered by the mask; it becomes darker and yellowish.
I'm a newbie and don't know how to fix this.