[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-72559":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":9,"language":10,"languages":9,"totalLinesOfCode":9,"stars":11,"forks":12,"watchers":13,"openIssues":14,"contributorsCount":15,"subscribersCount":15,"size":15,"stars1d":16,"stars7d":16,"stars30d":17,"stars90d":15,"forks30d":15,"starsTrendScore":18,"compositeScore":19,"rankGlobal":9,"rankLanguage":9,"license":20,"archived":21,"fork":21,"defaultBranch":22,"hasWiki":23,"hasPages":21,"topics":24,"createdAt":9,"pushedAt":9,"updatedAt":25,"readmeContent":26,"aiSummary":27,"trendingCount":15,"starSnapshotCount":15,"syncStatus":28,"lastSyncTime":29,"discoverSource":30},72559,"ComfyUI-LivePortraitKJ","kijai\u002FComfyUI-LivePortraitKJ","kijai","ComfyUI nodes for LivePortrait",null,"Python",2170,173,22,97,0,3,8,9,67.02,"MIT License",false,"main",true,[],"2026-06-12 04:01:06","# ComfyUI nodes to use [LivePortrait](https:\u002F\u002Fgithub.com\u002FKwaiVGI\u002FLivePortrait)\n## Update 2\n\nAdded another alternative face detector: https:\u002F\u002Fgithub.com\u002F1adrianb\u002Fface-alignment\n\n![image](https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F1a77752a-9688-4b6f-9363-736367ad711a)\n\nAs this can use blazeface back camera model (or SFD), it's far better for smaller faces than MediaPipe, that only can use the blazeface short -model.\nThe warmup on the first run when using this can take a long time, but subsequent runs are quick.\n\nExample detection using the blazeface_back_camera:\n\nhttps:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F40b1fdb4-0b1f-4ea8-8322-aa9151055db0\n\n## Update\n\nRework of almost the whole thing that's been in develop is now merged into main, this means old workflows will not work, but everything should be faster and there's lots of new features.\nFor legacy purposes the old main branch is moved to the legacy -branch\n\nChanges\n- Added MediaPipe as alternative to Insightface, everything should now be covered under MIT and Apache-2.0 licenses when using it.\n- Proper Vid2vid including smoothing algorhitm (thanks @melMass)\n- Improved speed and efficiency, allows for near realtime view even in Comfy (~80-100ms delay)\n- Restructured nodes for more options\n- Auto skipping frames with no face detected\n- Numerous other things I have forgotten about at this point, it's been a lot\n- Better Mac support on MPS (thanks @Grant-CP\n\nupdate to this update:\n- converted the landmark runner onnx model to torch model, not something I have done before and I didn't manage to do anything but make it .pth file, so you'll just have to trust me on it.\n  This allows running all this without even having onnxruntime, it runs on GPU and is about just as fast. It's available on the MediaPipe cropper node as option:\nWhen selected it's automatically downloaded from here: https:\u002F\u002Fhuggingface.co\u002FKijai\u002FLivePortrait_safetensors\u002Fblob\u002Fmain\u002Flandmark_model.pth\n\n![image](https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002Fc547f55a-9ef7-4bc7-85df-cdbab69a3ca8)\n\n\n# Examples:\n\nRealtime with webcam feed:\n\nhttps:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F31f77c10-b757-44ae-bb26-39e45ec0b2d9\n\nImage2vid:\n\nhttps:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002Fcfec0419-d1eb-4e67-8913-890eeb155eef\n\nVid2Vid: \n\nhttps:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F28438fcb-fbb0-4e4e-baf4-00fe06c455de\n\n\nI have converted all the pickle files to safetensors: https:\u002F\u002Fhuggingface.co\u002FKijai\u002FLivePortrait_safetensors\u002Ftree\u002Fmain\n\nThey go here (and are automatically downloaded if the folder is not present) `ComfyUI\u002Fmodels\u002Fliveportrait`\n\n# Face detectors\n\nYou can either use the original default Insightface, or Google's MediaPipe. \n\nBiggest difference is the license: Insightface is strictly for NON-COMMERCIAL use.\nMediaPipe is a bit worse at detection, and can't run on GPU in Windows, though it's much faster on CPU compared to Insightface\n\nInsightface is not automatically installed, if you wish to use it follow these instructions:\nIf you have a working compile environment, installing it can be as easy as:\n\n`pip install insightface`\n\nor for the portable version, in the ComfyUI_windows_portable -folder:\n\n`python_embeded\u002Fpython.exe -m pip install insightface`\n\nIf this fails (and it's likely), you can check the Troubleshooting part of the reactor node for alternative:\n\nhttps:\u002F\u002Fgithub.com\u002FGourieff\u002Fcomfyui-reactor-node\n\nFor insightface model, extract this to `ComfyUI\u002Fmodels\u002Finsightface\u002Fbuffalo_l`:\n\nhttps:\u002F\u002Fgithub.com\u002Fdeepinsight\u002Finsightface\u002Freleases\u002Fdownload\u002Fv0.7\u002Fbuffalo_l.zip\n\n*Please note that insightface license is non-commercial in nature.*\n","ComfyUI-LivePortraitKJ项目为LivePortrait提供了ComfyUI节点支持，使用户能够更便捷地进行实时面部捕捉与处理。其核心功能包括多种面部检测技术的选择（如Insightface和MediaPipe），以及改进的视频到视频转换算法，实现了接近实时的面部追踪与动画生成。特别值得注意的是，该项目通过优化计算流程提高了运行效率，并且对不同平台的支持更加友好，比如增强了Mac上的MPS兼容性。适用于需要实时或近实时面部动画的应用场景，如虚拟角色驱动、在线直播特效等。",2,"2026-06-11 03:42:34","high_star"]