You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, we've test to add an input image condition using the IP-Adapter-Plus (SDXL version). Following are some results and discussions.
In general, IPAdapter does have a style and content preservation, but we find it hard to balance the guidance strengths of different conditions (text & input-image) case-wisely.
Can we consider to also take the input image as condition (with e.g. ip-adapter) to realize personalized image generation ?
The text was updated successfully, but these errors were encountered: