If you are using gesture recognition or anomaly detection techniques for use cases such as predictive maintenance, here are some resources from Arm and partners that can help. Typically, anomaly detection uses various sensors, such as accelerometer, gyroscope, temperature or vibration. Go to section:
Gesture recognition with Arduino
Use gestures to train a classifier in TensorFlow Lite Micro and deploy it to the Cortex-M4 based Arduino Nano board running Mbed OS.View guide
Deploy a gesture recognition model with TensorFlow
This tutorial focuses on the use of TensorFlow Lite Micro on low-power microcontrollers to perform gesture recognition. The workshop covers ML model training and deployment of a gesture-recognition activity, “Magic Wand”.
Cartesiam.AI smart coffee machine
Build a device that sits directly on a coffee machine, learns its vibration patterns, and detects when coffee is ready. Create this device with an STM32 Nucleo-32 development board with the Cartesiam NanoEdge AI Library and Mbed OS.View guide
Cartesiam.AI smart vibration sensor
Create a device that sits on a vibrating machine, learns its vibration patterns, and detects potential abnormalities in its behavior. Create this device with an STM32 Nucleo-64 development board using Cartesiam’s NanoEdge AI Library, Pygame, and Mbed OS.
Edge Impulse motion recognition
Use an Arm Cortex-M4 based board with Edge Impulse to capture data and choose, train, and deploy a neural network model to detect different gestures. This guide also shows how to use classical ML techniques with standard ML techniques to detect motion inconsistencies.View guide
Predictive maintenance with TensorFlow Lite and Renesas
Learn how to integrate anomaly detection ML into a real-time motor application on the Renesas RA MCU family with TensorFlow Lite for Microcontrollers.Watch video
Get support with Arm Training courses. You can also open a support case or manage existing cases.