ComfyUI-LivePortrait_v2

https://github.com/user-attachments/assets/3bc8bb2a-7e2b-4c1f-9f09-55c38a7a86c7

https://github.com/user-attachments/assets/01ba6af3-9080-4af7-8795-11b8561e78f8

🔥 Updates

ComfyUI nodes for LivePortrait, We support animal image driven mode and regional control for Comfyui!!! We have developed animal expression-driven nodes for ComfyUI that have the same effect as the source code.

Introduction

This repo, named ComfyUI-LivePortrait_v2, thanks to paper LivePortrait: Efficient Portrait Animation with Stitching and Retargeting Control. We developed a custom_node for Liveportrait_v2 that enables flexible use on Comfyui to drive animal image-based emoji generation from videos.

Getting Started

Clone the code and prepare the environment

git clone https://github.com/VangengLab/ComfyUI-LivePortrait_v2.git
cd ComfyUI-LivePortrait_v2

In this node, we need dependencies related to XPose. Specifically, it needs to be configured and prepared according to the instructions on https://github.com/KwaiVGI/LivePortrait. The cuda version is preferably 12.1.

or you can refer to my environment on https://github.com/VangengLab/Comfyui_Liveportrait_v3/edit/main/README.md

Download pretrained weights

refer to https://github.com/VangengLab/Comfyui_Liveportrait_v3/edit/main/README.md this repo will tell you all details about pretrained weights

!!!Important reminder

Since the node we developed is relatively simple and we hope that users can get started quickly, we did not open a video upload window. We need to put the expression video (of a person) that needs to be driven into ComfyUI-LivePortrait_v2/assets/examples/driving and enter its name in the node. According to the example in the figure, it can be completed.