diff options
| author | hithereai <121192995+hithereai@users.noreply.github.com> | 2023-01-03 09:22:06 +0000 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2023-01-03 09:22:06 +0000 |
| commit | 9a3b0ee960b0c61c4f60e3081ae6f2098533d393 (patch) | |
| tree | efd4a016ddb479ab8bfde1f72b50bb919a0c7f0d /requirements.txt | |
| parent | fda1ed184381fdf8aa81be4f64e77787f3fac1b2 (diff) | |
| download | stable-diffusion-webui-gfx803-9a3b0ee960b0c61c4f60e3081ae6f2098533d393.tar.gz stable-diffusion-webui-gfx803-9a3b0ee960b0c61c4f60e3081ae6f2098533d393.tar.bz2 stable-diffusion-webui-gfx803-9a3b0ee960b0c61c4f60e3081ae6f2098533d393.zip | |
update req.txt
The old 'opencv-python' package is very limiting in terms of optical flow - so I propose a package change to 'opencv-contrib-python', which has more cv2.optflow methods.
These are needed for optical flow trickery in auto1111 and its extensions, and it cannot be installed by an extension as only a single package of opencv needs to be installed for optical flow to work properly. Change of the main one is Inevitable.
Diffstat (limited to 'requirements.txt')
| -rw-r--r-- | requirements.txt | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/requirements.txt b/requirements.txt index e2c3876b..4f09385f 100644 --- a/requirements.txt +++ b/requirements.txt @@ -9,7 +9,7 @@ gradio==3.15.0 invisible-watermark
numpy
omegaconf
-opencv-python
+opencv-contrib-python
requests
piexif
Pillow
|