My stuff

  • My Workflows

  • Liked Workflows

  • Following Workflows

Go to OpenArt main site
Upload workflow

audio-driven digital human animation (spark tts + latentsync)

5.0

0 reviews
4
2.3K
303
1
Description

https://github.com/1038lab/ComfyUI-SparkTTS


https://github.com/ShmuelRonen/ComfyUI-LatentSyncWrapper


Note: ShmuelRonen's ComfyUI-LatentSyncWrapper has serious optimization issues, and you need to make the following additional adjustments to run this node:
-- pip install pytorch_lightning
-- Modify line 226 of the file 'ComfyUI\custom_nodes\ComfyUI-LatentSyncWrapper\nodes.py' from: 'ffmpeg-python>=0.2.0' to 'ffmpeg-python'


Disadvantage: If you encounter the error 'Allocation on device' with the LatentSyncNode, please lower the resolution.


Due to the above-mentioned issues, LatentSync can only be used for experimentation and demonstration purposes, and practical applications are still a long way off.

Discussion

(No comments yet)

Loading...

Author

108
119.1K
873
430.0K

No reviews yet

  • - latest (9 months ago)

Primitive Nodes (3)

LatentSyncNode (1)

SparkTTS_AdvVoiceClone (1)

VideoLengthAdjuster (1)

Custom Nodes (6)

Comfy-RVC

  • - PreviewAudio (1)

ComfyUI

  • - SaveAnimatedWEBP (1)

  • - LoadImage (1)

  • - LoadAudio (1)

  • - VHS_VideoCombine (1)

  • - ImageResizeKJ (1)

Checkpoints (0)

LoRAs (0)