Arm NN is a neural network inference engine developed by Arm and the Linaro Machine Learning initiative connecting popular frameworks such as TensorFlow, TF Lite, ONNX, or Caffe with the CPU, GPU or NPU on your device using Arm NN’s backends. The new Python interface called PyArmNN makes this even easier by enabling Python and all its modules.

The presentation takes you through the concepts of the Arm NN framework, dives deeper both into Python enablement, the usage, and into the different backends for acceleration on platforms such as Arm NEON, OpenCL, or having to develop a custom one.