Be aware that this repository contains compiled tensorflow model files.
Tensorflow model files are programs which can perform arbitrary operations on your computer.
More details here: https://github.com/tensorflow/tensorflow/blob/master/SECURITY.md
If you are running any branch or fork of Cacophony Project's audio-analysis code, be sure to confirm the origin of any tensorflow model you run, and take precautions accordingly..
Build the docker image
docker build -t cacophony-audio .
Then run it with the parameters:
- model_dir - model base directory
- base_dir - base directory
- audio-file - audio file relative to base_dir
- metadata-file- meta file relative to base_dir (not used at the moment)
- path to models- path to models directory contains (model1, model2, and model3) relative to model_dir
docker run -it-v {model_dir}:/model -v {base_dir}:/io cacophony-audio /io/{audio-file} /io{metadata-file} /model/"."
Make a tag on GitHub This will push a docker image to docker.io to use this image update cacophony-processing tag version To run on docker image use the following command