The Linux foundation AI Project (LF AI), an organization created to establish an ecosystem for open-source innovation in the fields of artificial intelligence (AI), machine learning and deep learning, recently announced NNStreamer as its latest Incubation Project. An open-source set of GStreamer plugins developed and released by Samsung Electronics, NNStreamer offers a simple means for GStreamer developers to adopt neural network models, and for neural network developers to manage both neural network pipelines and their filters.

NNStreamer makes developing on-device AI systems easier and more efficient. It does this by enabling the description of general systems with various inputs, outputs and processors, and neural networks with the Pipe and Filter architecture. NNStreamer provides easy-to-use APIs with corresponding SDKs spanning C APIs (for all platforms), Tizen .NET (C#), and Android (Java), as well as a wide range of neural network frameworks and software platforms including Ubuntu, macOS and OpenEmbedded. It became open source in 2018, and is currently under active development through both the Tizen project and via a wide range of consumer electronic devices.

NNStreamer’s selection as an Incubation Project grants it access to the wide range of benefits that LF AI’s support entails. As LF AI’s Executive Director, Dr. Ibrahim Haddad, explained, the foundation is especially excited to welcome NNStreamer to the fold.

“We are very pleased to welcome NNStreamer to LF AI,” said Dr. Haddad. “Machine learning applications often process online stream input data in real time, which can create a complex system. NNStreamer can be used to easily represent and efficiently execute against these challenges. We look forward to supporting this project and helping it thrive under neutral, vendor-free and open governance.”

Additional information on NNStreamer may be found here.