This is a workflow for swapping a character's head from a reference image into a video. It works for both realistic and anime-style reference images. For important details, please see the testing and usage instructions below.
馃捇I've already set up an 鉃★笍 AI application and the complete 鉃★笍 workflow for you. This will let you quickly try it out online and avoid the hassle of not having enough VRAM or dealing with local plugin conflicts.
馃巵Bonus: If you're signing up for the first time, you can get 1,000 free RH Coins by using my link and the invite code 鉃★笍 rh-v1182. Plus, you'll get another 100 RH Coins for logging in daily.
馃殌Workflow Testing and Usage Instructions:
1. This model is focused only on swapping the character's head. If you want to replace the entire body and clothing, you can use this other workflow.
2. You don't have to write a prompt, but if you want to get some other different effects, you can add a detailed description in the prompt.
Description
FAQ
Comments (19)
HI :) where i can find TT_img_enc? comfyUI model manager doesn't find it, and tt_tools's git it's down
Replace it with a simple VHS video Combine ;) should be working
Puoi cancellarlo e sostituirlo con un output video standard. Quel nodo serve solo per codificare immagini NSFW generate su piattaforme online (come runninghub), perch猫 altrimenti i filtri non permettono di salvarle
@EKKIVOK聽where do I get that from plz?
@jonog247634聽you have already that node..... just type in search by clicking with right click on your workflow :)
@EKKIVOK聽Ah yep thanks :)
Where can I find the Mocha model?
Hi, thanks! It's probably because I'm not familiar with the interface and nodes, but the VRAM problem just makes it unusable for me, even with very low resolution images and very short videos.
Hi, how did you install mochaembeds node ?
It's now in the ComfyUI-WanVideoWrapper custom nodes but it didn't show up when I did an update to this set of nodes. I had to uninstall the ComfyUI-WanVideoWrapper then reinstall it with a nightly build... then it shows up
although my current GPU is 24g with RTX4090, but it still show OOM. not sure how we can fine tune the workflow to release more memory
Enable block swap but turn 'use_non blocking' to false & enable VAE tiling. I got it to work with my 5070ti with 64gb memory - at 720x480 takes 3-5mins.
Why
Error in image processing: Error in batch processing: 'numpy.float64' object has no attribute 'cpu'
Missing nodes, install missing nodes, restart, missing nodes, install missing nodes, restart, missing nodes,install missing nodes, restart, missing nodes,install missing nodes, restart, missing nodes,install missing nodes, restart, missing nodes,install missing nodes, restart, missing nodes,install missing nodes, restart, missing nodes,install missing nodes, restart, missing nodes,
I have a 4090, but it still said I am out of memory for my GPU, bruh
anyone actually get this to work? I keep getting error after it finished the mediapipe ,installed mediapipe then I get a operands could not be broadcast together with shapes (832, 480, 4) error
I don't think this was designed to work, as nobody has posted any videos, and it's more likely a ploy to get people to click his online runpod affiliate link