Edge Impulse Inc. / Tutorial: Continuous motion recognition

This is the finished Edge Impulse project for the tutorial 'Continuous motion recognition'. From here you can acquire new training data, design impulses and train models.

About this project

Classify Your Gestures

Snake motion gif

Have you ever wondered how you can use acceleration to predict the type of input motion? Clone this project to build an embedded ML project to detect various hand gestures from your device's accelerometer!

You can also follow our tutorial to guide you through building your continuous motion recognition model, from data collection to deployment on embedded devices.

Sensor & Block Information

  • Accelerometer data (.cbor files) @ 62.5 Hz
  • Spectral Features DSP block for time-based sensor data
  • Neural Network Classifier with prediction outputs: "idle", "snake", "updown", "wave"
  • Anomaly Detection block for anomaly score output of unknown motions (gestures/motions the model was not trained on)

Creating your first impulse (100% complete)

Acquire data

Every Machine Learning project starts with data. You can capture data from a development board or your phone, or import data you already collected.

Design an impulse

Teach the model to interpret previously unseen data, based on historical data. Use this to categorize new data, or to find anomalies in sensor readings.

Deploy

Package the complete impulse up, from signal processing code to trained model, and deploy it on your device. This ensures that the impulse runs with low latency and without requiring a network connection.

Download block output

Title Type Size
Spectral features training data NPY file 7522 windows
Spectral features training labels NPY file 7522 windows
Spectral features testing data NPY file 1552 windows
Spectral features testing labels NPY file 1552 windows
NN Classifier model TensorFlow Lite (float32) 6 KB
NN Classifier model TensorFlow Lite (int8 quantized) 4 KB
NN Classifier model TensorFlow SavedModel 17 KB
Anomaly detection model JSON 7 KB

Summary

Data collected
15m 26s

Project info

Project ID 76311
Project version 9