Neural Network Support as Gstreamer Plugins.
NNStreamer is a set of Gstreamer plugins, which allows Gstreamer developers to adopt neural network models easily and efficiently and neural network developers to manage stream pipelines and their filters easily and efficiently.
Architectural Description (WIP)
-
Provide neural network framework connectivities (e.g., tensorflow, caffe) for gstreamer streams.
- Efficient Streaming for AI Projects: Neural network models wanted to use efficient and flexible streaming management as well.
- Intelligent Media Filters!: Use a neural network model as a media filter / converter.
- Composite Models!: Allow to use multiple neural network models in a single stream instance.
- Multi Modal Intelligence!: Allow to use multiple sources for neural network models.
-
Provide easy methods to construct media streams with neural network models using the de-facto-standard media stream framework, GStreamer.
- Allow any gstreamer users to put neural network models as if they are media filters.
- Allow any neural network developers to manage media streams fairly easily.
Note that this project has just started and many of the components are in design phase. In Component Description page, we describe nnstreamer components of the following three categories: data type definitions, gstreamer elements (plugins), and other misc components.
For more details, please access the following manual.
- Press Here
- Example app with camera (stable)
- Wiki page usage example screenshots (stable)