Human Activity Recognition Example

This example implements a simple HAR model. It collects data from an IMU accelerometer and gyroscopic sensor, and classifies the data into one of 5 activities (sitting, standing, walking, jogging, go up stairs). This example is based on the Human Activity Recognition Model Zoo repository.

The model classifies the last 10 seconds of movement every 2s.

To use this model, you'll need to connect the sensor (an MPU6050) to the EVB (instructions here).

Required Hardware and Configuration

This example is requires the following hardware: * Apollo510 EVB * ICM-45605 Mikroe Board (MIKRO-6040) installed in EVB's Clickboard slot * J13 jumper selecting 3.3v

Building and Running the Example

This is the primary neuralSPOT example, so the makefile defaults many of its options to it.

$> cd .../neuralSPOT
$> make clean
$> make -j EXAMPLE=ai/har deploy
$> make view # connects to EVB SWO via SEGGER Jlink

Additional Example Code

There are 2 additional files (*.dnc, which stands for 'do not compile') with alternate implementations: 1. har.cc.ap4: uses an i2c-based IMU instead of the SPI ICM IMU 2. har.cc.blockmode: collects 10s blocks when a button is pressed instead of using a continuous window