Train gesture models on the IdeaBoard
Disconnected

1 Data

Add a class for each gesture, then hold the board still in the pose / motion and record samples.

Live IMU

Accelerometer (m/s²)
x: –   y: –   z: –
Gyroscope (°/s)
x: –   y: –   z: –

2 Train

Record at least 2 classes with a few samples each to enable training.

3 Test (live)

Train a model to see live predictions.

4 Export to IdeaBoard

Generates a standalone code.py that runs the trained model on-device.

Generated CircuitPython
// Train a model first.