Replies: 5 comments
-
@brunogcar Thanks a lot |
Beta Was this translation helpful? Give feedback.
-
Tried to update this mapping for ControlNet 1.1: model -> preprocessor
preprocessor -> model
What about mediapipe_face? scribble_xdog? threshold? @lllyasviel could you check if this makes sense and if anything is missing? |
Beta Was this translation helpful? Give feedback.
-
since 1.1, cn begin to use standard naming, and memorizing correspondence is not necessary now. if model and preprocessor match a full word, they match. otherwise, they do not match. if no preprocessor matches, use none. threshold is for advanced users who understand what is happening and sometimes they want to threshold photos of scribbles drawn on paper. note that now t2i adapter also follow the word matching rule and it is not necessary to remember matches. |
Beta Was this translation helpful? Give feedback.
-
What's the difference between softedge_hed, softedge_hedsafe, softedge_pidinet and softedge_pidisafe? |
Beta Was this translation helpful? Give feedback.
-
i tried to compile a list of models recommended for each preprocessor, to include in a pull request im preparing and a wiki i plan to help expand for controlnet
some are obvious, but others aren't that much, specially to new users
canny -> control_canny - t2iadapter_canny
mlsd -> control_mlsd
hed -> control_hed
scribble -> control_scribble - t2iadapter_sketch
fake_scribble -> control_scribble - t2iadapter_sketch
openpose -> control_openpose - t2iadapter_openpose - t2iadapter_keypose
openpose_hand -> control_openpose - t2iadapter_openpose
segmentation -> control_seg - t2iadapter_seg
depth -> control_depth - t2iadapter_depth
depth_leres -> control_depth - t2iadapter_depth
depth_leres_boost -> control_depth - t2iadapter_depth
normal_map -> control_normal
binary -> control_scribble - t2iadapter_sketch
color -> t2iadapter_color
pidinet -> control_hed
clip_vision -> t2iadapter_style
please let me know if i missed something or got it wrong, idea isnt to make a list of everything that woks, but which models are ideal for each preprocessor
Beta Was this translation helpful? Give feedback.
All reactions