References
Setup
Get the sources
Install Visual C++ Tools for Cmake
Install the CUDA Toolkit 10.1
Install the NVIDIA CUDA Deep Neural Network library (cuDNN v7.6.5 (November 5th, 2019), for CUDA 10.1)
Install and configure Python
Here are some references that helped me to build this experiment:
- Hands-On Computer Vision with TensorFlow 2: Leverage deep learning to create powerful image processing apps with TensorFlow 2.0 and Keras by Benjamin Planche and Eliot Andres
- YoloV3 Implemented in TensorFlow 2.0 by Zihao Zhang
- Face Recognition
You may
- clone this github repository
- or download a zip containing the latest version or a given release of the code
- Install Visual Studio build tools from here.
- In Visual Studio 2017 go to the Individual Components tab, Visual C++ Tools for Cmake, and check the checkbox under the "Compilers, build tools and runtimes" section.
- Please install from the following link
Note that you can uncheck "Install Visual Studio Extensions" in the options.
Install the NVIDIA CUDA Deep Neural Network library (cuDNN v7.6.5 (November 5th, 2019), for CUDA 10.1)
- Follow the instructions detailed here
Make sure to install the version v7.6.5 (November 5th, 2019), for CUDA 10.1
- Install Python 3.7 or later.
- From the root folder of the project, type
pip install virtualenv
- Then type
This will create a venv subfolder.
virtualenv --python="C:\Users\[YOUR USER NAME]\AppData\Local\Programs\Python\Python37\python.exe" venv
Note that the path to the python.exe (v3.7) may vary on your machine. If python has been installed with visual studio, you may have something similar to this:virtualenv --python="C:\Program Files (x86)\Microsoft Visual Studio\Shared\Python37_64\python.exe" venv
- From the root folder of the project, activate the virtual environment by typing:
.\venv\Scripts\activate.bat
- Install packages:
pip install -r requirements.txt