Extensions/ComfyUI-EffectErase
ComfyUI Extension

ComfyUI-EffectErase

A ComfyUI extension with 2 custom nodes.

By alisson-anjos·Created 5 months ago·Updated 5 months ago· 6
alisson-anjos/ComfyUI-EffectErase
Nodes2
On cloudLocal install
CategoryEffectErase, EffectErase/Utils
Stars6
Updated5 months ago
Readme

ComfyUI-EffectErase

A custom node for ComfyUI that implements Joint Video Object Removal and Insertion using the Wan2.1 1.3B Video DiT architecture.

This repository allows you to perform advanced, temporally consistent Video-to-Video inpainting and object removal directly inside your ComfyUI workflows without complicated setups.

Features

  • Object Removal: Erase moving objects from videos by providing an input video and a mask.
  • Auto-Dependencies: Automatically installs modelscope, peft, and HuggingFace dependencies at startup.
  • Model Auto-Download: Automatically fetches the required Wan2.1 model and EffectErase LoRAs on the first generation and saves them to models/EffectErase/.
  • Acceleration LoRA Support: Supports attaching native Wan2.1 acceleration LoRAs to speed up inference times.
  • FlowMatch Scheduler Configuration: Exposes sigma_shift, CFG, inference steps, and seed controls.

Usage

Simply plug the EffectEraseObjectRemoval node into your ComfyUI workflow.

  1. Provide a video_fg_bg (Image Tensor representing video frames).
  2. Provide a video_mask (Target Object Mask).
  3. Connect the output to a Save Video or Video Combine node.

Acknowledgements & Credits

This node is just a wrapper to bring the state-of-the-art research into the ComfyUI ecosystem. All credit for the mathematical implementation, training, and neural architecture modifications belongs entirely to the original authors.

  • EffectErase: Developed by FudanCVL. Official Repository
  • DiffSynth-Studio: The inference framework used to execute the custom modified WanRemovePipeline natively. DiffSynth-Studio
  • Wan2.1: The underlying Video Diffusion Transformer baseline.

If you use this node in your research or projects, please consider starring and citing the original EffectErase repository.