Create a new env using this command mkvirtualenv "name_env" -p python3 change "name_env" for your choice name.
Enter inside env: workon "name_env" .
Install dependences: pip -r requirements.txt.
Clone this respository using git clone https://github.com/ximenesfel/action_recognition_openpose.git command.
Usage
To write json files from videos (Recommended: GPU)
Open write_json_from_video.py and fill this variables: path_videos (path to your videos folder) and path_output_data (path where you like to save json data). Please insert all path. Ex: /home/"user"/Desktop/.....
Run python write_json_from_videos.py.
Create database from json files (Recommended: GPU or CPU)
Open generate_data.py and fill the variable path_data (path where you save json files). Please insert all path. Ex: /home/"user"/Desktop/.....
Run python generate_data.py.
The file database.dat will be saved in your current directory with all data from json's files.
Read data form database (Recommended: GPU or CPU)
Open read_data.py and fill the variable data_path (path where you save database file).