Simple video2video workflow

详情

模型描述

Inputs :

  • Frames

  • Processing resolution

  • Seed

  • Steps

  • CFG

  • Model and loras

  • Video to imitate

You can also tweak :

  • Skip frames to select a specific part of the input video

  • K Sampler algorithm and details

  • Add postprocessing such as bigger upscaling, face detailer etc.

  • Play around with the Control Net weights ; 0 depth and 0 lineart is better for some cases

  • Add a motions lora

Don't forget to install the custom nodes with ComfyUI manager.

此模型生成的图像