david-svitov / baldgan Goto Github PK

Should be:

tensorflow==2.11.0
retina-face==0.0.13

Can I resume the training with custom data that have a portrait image and is there is additional step I need to perform because I am not getting good results?

Hello, what versions of tensorflow and keras were used for this project?

I have an interest in this project. So I tried to test this code but failed to install opencv-python and keras-contrib.
Currently having this kind of error message
Enviroment :
OS : Ubuntu 20.04LTS
Python : 3.9
Conda : MiniConda 4.10

Solving environment: failed with initial frozen solve. Retrying with flexible solve.

PackagesNotFoundError: The following packages are not available from current channels:

  - opencv-python==4.5.1.48
  - //www.github.com/keras-team/keras-contrib.git

Current channels:

  - https://conda.anaconda.org/conda-forge/linux-64
  - https://conda.anaconda.org/conda-forge/noarch

To search for alternate channels that may provide the conda package you're
looking for, navigate to

    https://anaconda.org

and use the search bar at the top of the page.

1. The requirements.txt is not enough to satisfy from retinaface import RetinaFace.
2. When I pip install retinaface from here, the method model.detect() is not supported (model.predict is) so I guess this is not the right library. I also tried installing all the other versions of this library and they did not have this method.

Question is: which library enables this import ?