Face Recognition Baseline

This repo is a starting point for using Face Recognition Library

Dependencies

• Python>=3.6
• CUDA=10.0
• MXNet

Installation

Make sure conda is installed.

Note that: If you are using GPU, you need to change the version of cuda in requirements.txt file.

mxnet-cu101 means the package is built with CUDA/cuDNN and the CUDA version is 10.1.

# create environment 
conda create --name face_recog_test python=3.6
conda activate face_recog_test

# install dependencies for cpu 
pip install -r requirements-cpu.txt
# or install dependencies for gpu 
pip install -r requirements-gpu.txt

# install face recognition library
pip install --upgrade greenlab-library

Getting Started

Steps:

1. Change the image path in and parameters in configs/server_api.yaml file.
2. Run main.py file
3. Check the results at logs/outputs/test.png