Real-time pipeline for video analytics.
-
pipert: The library, which contains:
- core: The core parts of the library, the pipeline base 'component' and component's base 'routine'.
- utils: general purpose functions for monitoring, controlling, and massage passing between components and routines.
-
pipert.contrib: The Contrib directory contains implemented components and routines for various projects.
The code in pipert.contrib is not as fully maintained as the core part of the library. It may change or be removed at any time without notice.
API documentation and an overview of the library can be found here.
- In order to use the submodules in this project do the following::
-
When cloning:
git clone --recurse-submodules https://github.com/ItamarWilf/PipeRT.gitOR
-
After
git pull:git submodule update --init --recursive
-
-
Docker
-
Docker compose
-
Nvidia docker: https://github.com/NVIDIA/nvidia-docker
-
nvidia-container-runtime: https://github.com/NVIDIA/nvidia-container-runtime
-
Running the pipeline for the first time:
- docker-compose up -d
-
Build the pipeline after changes:
- docker-compose up -d --build
-
After the pipeline is up run the cli.py script and connect (The default endpoint is tcp://0.0.0.0:4002)