Setting up ComfyUI workflow: transferring face and body to video (Wan 2.2 / Animate)
Looking for a specialist in ComfyUI with experience in video generation based on Wan 2.2 / Animate.
Task: to assemble and configure a ready-made workflow for overlaying a person's appearance from a photo onto a video. Three modes of operation are needed:
1. Face + hair - high-quality transfer of the face while preserving the hairstyle, head shape, and hairline (not just an oval face).
2. Full body - transfer of the entire figure with correct anatomy, without artifacts on the arms and legs.
3. Finger retouching - a separate pass or built-in correction to eliminate typical artifacts of the hands
Technical requirements:
- Video generation base: Wan 2.2 or Animate
• For appearance transfer, use LoRA, IP-Adapter, ControlNet, or similar solutions of your choice, the main thing is the quality of the result
- Launch on RunPod via cloud GPU
- Submission format: workflow with deployment instructions, or a fully configured environment, at your discretion
Testing:
Before acceptance, you need to run our test video and photo through the proposed setup and show the result. We will provide the materials, you run it on RunPod and demonstrate the operation of all three modes
If you have a ready or proven setup for this task - write, show examples
-
633 2 0 Hello. I am ready to take on the implementation of this pipeline on Wan 2.2. Regarding the technical part, there are important clarifications:
1. About RunPod: I am developing and testing the workflow in the ComfyUI environment. I will provide you with a ready .json file with a complete map of custom nodes and setup instructions. Deployment directly on cloud resources is a separate infrastructure process, so the technical part of the deployment (Server/Docker) should be handled by your administrator.
2. About hand correction: It is important to understand that in AI video generation at this stage of technology development, there is no "automatic" perfect generation of fingers without artifacts in motion. I will implement an advanced pipeline using masking through SAM2 for local Inpainting. This will yield the best possible result, but be prepared that micro-correction of masks will be needed for complex scenes. However, I do not guarantee that it will be 100% perfect.
I made a sample video https://www.youtube.com/watch?v=eoOI_eV0bP0
If this approach (architectural honesty + professional tools) suits you — we can get started.
-
Це називаєтся діпфейк. Але якщо навіть надасте письмову згоду що це буде ваше обличчя. То потрібна ціна. Якщо мене зацікавить цифра в оголошенні я обговорюватиму це питання
-
Дорисовка пальцев - отдельный проход или встроенная коррекция для устранения типичных артефактов кистей рук- не слышал про такое,это вы придкмали или где то есть инфа,источник есть,делал подобный проект без развертывания,но насчет пальцев это фантастика.
-
Current freelance projects in the category Video Creation by Artificial Intelligence
Editing and voiceoverGood day! I need help with video editing and sound design. Overall, the video has no dialogues. Just specific sounds. A bit of music somewhere. The video is about 1-1.5 minutes long. Video Creation by Artificial Intelligence, Audio & Video Editing ∙ 3 hours 46 minutes back ∙ 23 proposals |
Installation
22 USD
Looking for a video editor to create a dynamic clip. What needs to be done: • edit the video from the footage of a girl dancing; • create a modern, dynamic edit in the style of trendy Reels/TikTok; • select music that highlights the dynamics well; • add smooth and impressive… Video Creation by Artificial Intelligence, Audio & Video Editing ∙ 6 hours 26 minutes back ∙ 40 proposals |
Animate a photo (series of 10 photos)Hello! I need a specialist who understands AI. I need to animate about 10 photos (a birthday gift) and create a mini video. Please suggest options and a price for the work. Thank you! Video Creation by Artificial Intelligence, Photo Processing ∙ 7 hours 3 minutes back ∙ 29 proposals |
Generation of a one-minute video in a neural network.Generation of a one-minute video in a neural network. The video needs to be created based on the script that I will provide for advertising a medical center. Video Creation by Artificial Intelligence ∙ 9 hours 55 minutes back ∙ 25 proposals |
Создание креативного motion-видео
25 USD
Project Description We are looking for an experiencedMotion Designer to create a premium-quality animated video for ourArt of Joy campaign. The video will be displayed on large LED screens during events, exhibitions, and brand activations.Concept The main objective is to convey… Video Creation by Artificial Intelligence, Audio & Video Editing ∙ 15 hours 3 minutes back ∙ 12 proposals |