Installation of Face_Recognition on python 3.6 does not recognize that dlib has already been installed in the system.

conda install dlib scipy numpy successfullypip install face_recognition and the installation tried to reinstall dlib again.further investigation revealed that pip install face_recognition checked for info in /opt/conda/envs/py36/lib/python3.6/site-packages which seems to only contain dlib.so and without related *.egg-info or *.dist-info file.

I installed dlib from channel conda-forge , anybody know why is this missing?
Sounds like a potential problem with the conda package. You should probably file an issue on https://github.com/conda-forge/dlib-feedstock and ask @patricksnape. I unfortunately don't know anything about how conda-forge works.
I guess in the meantime you could hack around it with pip install --no-dependencies face_recognition face_recognition_models
That's just how dlib builds. The new(ish) pip installs of dlib are not currently compatible with conda-forge since we need patches to even get dlib to build. Pip isn't so great at detecting conda packages so you'd be better off making all the deps conda packages if you really want full compatibility.
using --no-dependencies solved the problem for me. Installing dlib manually is a great hassle and conda with conda-forge channel solved that problem. I strongly recommend you include this option in setup guide because it is a lot more accessible compare to downloading boost and compile dlib by hand.
hello guys,
I have installed anconda 5.3.0(32bit) from https://www.anaconda.com/download/#windows
Also opencv(2.4.13) and python 2.7 are already installed in my laptop. Now, i want to install dlib library for face recognition.
Q1. Is it possible to only install dlib using anaconda leaving other installations as it is?
Q2. I even want to download the upgraded version of openCV (3.2.x), will it cause any problem in the setup of other things?
Please don't ask new questions on old closed issues. No one is going to see them.
Most helpful comment
I guess in the meantime you could hack around it with
pip install --no-dependencies face_recognition face_recognition_models