Arm provides a range of software solutions which seamlessly plug into existing ML frameworks, allowing developers get the very best performance from their underlying Arm hardware.

The following guides are designed to provide a quick on-boarding to the Arm software solution. The guides demonstrate the performance gain that can be achieved when utilizing software that is specifically optimized for Arm based devices. The guides cover several ML use cases across a range of Arm IP including Cortex A CPU, Cortex M CPU, and Mali GPU. Each guide specifies the environment setup, provides example source code, and reference neural networks.

Whether you are developing for Android, Raspberry Pi, Odroid-N2, or a low-powered embedded solution, our guides are a useful introduction to get started with Arm ML software solutions.

Black coloured computer vision icon

Object detection

Real time object detection with YOLOv3 using PyArmNN and Debian Packages. 

View the guide
Black coloured voice icon

Speech recognition

Perform Automatic Speech Recognition (ASR) on Raspberry Pi with wav2letter using Arm NN and Debian Packages

View the guide
Image classification icon

Image classification

Identify and classify people, objects, or animals using MobileNet V2 and Arm NN 

View the guide
Sensor icon

Key word detection

Build an AI voice assistant with Tensorflow Lite for microcontrollers, Mbed OS and a Cortex-M based MCU.

View the guide

Other guides