Design Win

STM32Cube function pack for ultra-low power context awareness with distributed artificial intelligence (AI): acoustic analysis with NN on MCU and motion analysis with ML on IMU

Get Software Download databrief

Product overview


FP-AI-CTXAWARE1 is an STM32Cube function pack featuring examples that let you connect your context awareness node to a smartphone via BLE and use a suitable Android™ or iOS™ application, like the STBLESensor app, to configure the device.

The package enables advanced applications such as human activity recognition (HAR) or acoustic scene classification (ASC), on the basis of outputs generated by the LSM6DSOX machine learning core (MLC) for HAR and the neural networks (NN) for ASC running on the STM32L4R9ZIJ6 MCU. The machine learning for HAR is a decision tree logic algorithm generated by Unico-GUI. The NN are implemented by a multi-network library supporting both floating and fixed point arithmetic, generated by the X-CUBE-AI extension for STM32CubeMX tool. The NN provided in this package are just examples of what can be achieved by combining the output of X-CUBE-AI with connectivity and sensing components from ST.

This package, together with the suggested combination of STM32 and ST sensors, can be used to develop specific wearable AI applications, where ultra-low power consumption is a key requirement, thanks to distributed deep edge AI approach.

The software runs on the STM32 microcontroller and includes all the necessary drivers for the STEVAL-MKSBOX1V1 evaluation board.

  • All features

    • Complete firmware to develop a context awareness node with BLE connectivity, digital microphone, environmental and motion sensors, performing real-time monitoring of sensors and audio data
    • Machine Learning Core (MLC) featuring real-time human activity recognition (HAR) generated thanks to Unico-GUI and running on LSM6DSOX
    • Middleware library generated thanks to STM32CubeMX extension called X-CUBE-AI, featuring example implementation of neural networks for acoustic scene classification (ASC) application
    • Multi-network support: concurrent execution of the MLC for HAR and the neural network for ASC
    • Ultra-low power implementation based on the use of an RTOS
    • Compatible with STBLESensor application for Android/iOS, to perform sensor data reading, audio and motion algorithm feature demo in standalone or combined views, and firmware update over the air (full FOTA)
    • Sample implementation available for STEVAL-MKSBOX1V1 evaluation board
    • Easy portability across different MCU families, thanks to STM32Cube
    • Free, user-friendly license terms

Get Software