There are two main implementations of hrnet
(1) official implementation
(2) openmmlab
1. Code implementation
1.1 hrnet official code
The hrnet official code does not implement the detection code, and does not provide a detection frame, which is not convenient overall
1.1.1 Environment error report
OSError: Could not find library geos_c or load any of its variants ['libgeos_c.so.1', 'libgeos_c.so']
in linux system
sudo apt-get install libgeos-dev
1.2.1 Environment configuration
Enter the folder where hrnet is located, or go to the created virtual environment
sudo apt-get install libgeos-dev
pip install -r requirements.txt
#Install COCOAPI:
# COCOAPI=/path/to/clone/cocoapi
git clone https://github.com/cocodataset/cocoapi.git $COCOAPI
cd $COCOAPI/PythonAPI
# Install into global site-packages
make install
# Alternatively, if you do not have permissions or prefer
# not to install the COCO API into global site-packages
python3 setup.py install --user
1.2 openmmlab implementation
For the environment configuration, you can refer to the openpose installation document , and I won’t write in detail.
1.2.1. Error reporting
(1) The pytorch version does not correspond
The latest version is installed directly, and the version does not correspond
2 Introduction to .hrnet
【Reference】Introduction