Now that you have things working, here are a few new things to try.
-
Add some more gestures to go with the 👊 and :flex:
- Load the IMU_Capture sketch on your Arduino Nano 33 BLE
- Caputure some additional gestures for a new emoji. Perhaps a 😄, 👍, 👏, or 👋
- Load the new CSV files into your Colab model
- Put the trained model back on your Nano 33 BLE
-
Try increasing and decreasing the number of recordings per gesture, how does this impact performance?
-
Try to only use the accelerometer or gyroscope data (not both), how does this impact performance?
-
Tweak the model structure and parameters
- Can you get better results?
- Can you reduce the size and still get "good" results
-
Check out the examples in the TensorFlow Lite Getting Started Guide
- Grab the TinyML book to continue at home