【Wan VACE】Video Inpaint Change Clothes


Updated:

Workflow Preview

Showcases (Image/Video)

上傳影片素材,自己擷取影片第一幀去準備好換成想要的服裝以作為參考圖,之後在Segment Prompt填入想要inpaint的區域

Upload your video. Manually extract the first frame of the video to prepare as a reference image with the desired clothing/outfit for inpainting. Afterward, fill in the desired inpainting region in the Segment Prompt.

Nodes Detail

48 Nodes
Primitive Node Types
10
  • LoraLoader3
  • PreviewImage1
  • KSampler1
  • VAELoader1
  • CLIPLoader1
  • CLIPTextEncode2
  • VAEDecode1
Custom Node Types
38
  • easy imageRemBg1
  • ModelSamplingSD31
  • FluxGuidance1
  • ImageCompositeMasked1
  • TrimVideoLatent1
  • WanVideoBlockSwap1
  • WanVideoModelLoader1
  • GrowMask1
  • BlurMaskFast1
  • Note2
  • TensorArt_PromptText2
  • EmptyImage1
  • GroundingDinoModelLoader (segment anything)1
  • ImageResizeKJv22
  • SAMLoader1
  • WanVideoVACEModelSelect1
  • UNETLoader1
  • TensorArt_LoadVideo1
  • VHS_SelectEveryNthImage1
  • GroundingDinoSAMSegment (segment anything)1
  • TensorArt_SelectParams2
  • Int1
  • MathExpression|pysssss1
  • WanVaceToVideo1
  • TensorArt_SelectResolution1
  • math_calculate2
  • StringToInt2
  • RIFE VFI1
  • VHS_VideoCombine2
  • TensorArt_SelectParamsInteger1
  • TensorArt_LoadImage1