This is the WorkFlow for ConsistencyLoRAs. It is a just basic Wan2.2 I2V workflow actually. The image you input is the object you want to keep consistent. For example, a car for the CarConsistency , a clothing for the ClothingConsistency. And before you using the LoRA, make sure you read the red lines in the model description page , make sure you use it in the right way.
Description
FAQ
Comments (2)
does this work for faces as well?
I doesn't train with face, but some sample posted by other users, it fit with other T2V character lora.