ORBIS

Appearance

Train your own wake word ​

ORBIS detects wake words with openWakeWord — small, on-device models that run before the main speech pipeline. The picker in Settings → Activation ships "Hey Orbis" plus the stock openWakeWord set, but you can train a model for any phrase you like ("Hey Computer", your own name, …) and drop it in.

A wake model is tiny (~200 KB–1.5 MB) and trains on synthetic speech, so you don't record anything yourself — a free Colab GPU does it in under an hour.

1. Train the model ​

openWakeWord ships an automatic training pipeline. The easiest path is its Colab notebook:

  1. Open openWakeWord → automatic model training (the automatic_model_training.ipynb notebook — "Open in Colab").
  2. Set your target phrase (e.g. hey computer). Keep it 2–4 syllables and distinct from everyday speech — short or common phrases false-fire.
  3. Run the cells. It synthesizes thousands of positive samples (via Piper TTS) plus hard negatives, trains a small classifier, and exports an .onnx.
  4. Download your_phrase.onnx when it finishes (~30–60 min on the free GPU).

Want the exact recipe we used for "Hey Orbis"? It's at protoLabsAI/hey-orbis-wakeword.

2. Install it in ORBIS ​

Your model rides on two shared front-end models (mel-spectrogram + embedding) that ORBIS already downloads for any wake word — you only add your classifier. Drop the file into the wake-word models directory:

~/Library/Application Support/studio.protolabs.orbis/models/wakeword/

Put your_phrase.onnx there next to melspectrogram.onnx and embedding_model.onnx (download any built-in wake word once in Settings → Activation to fetch those two if you haven't). The file's name (minus .onnx) is its id, and ORBIS turns it into a display phrase — hey_computer.onnx → "Hey Computer".

3. Select + tune it ​

  1. Settings → Activation → choose the Wake word style and select your model from the list.
  2. Relaunch — activation settings apply on the next launch.
  3. Say your phrase. The orb's status pill shows it while armed and flips to listening when it fires.
  4. If it's unreliable, tune the Sensitivity slider (lower = fires more easily) and relaunch. A clean utterance should score far above a background of ~0.

Troubleshooting ​

  • Never fires. First check your mic input level — System Settings → Sound → Input → raise Input volume. openWakeWord needs a real signal; a near-silent mic (the meter barely moving) won't trigger it no matter the threshold. See Voice isn't working.
  • False fires. Raise the sensitivity threshold, or pick a longer / more distinctive phrase and retrain.
  • Won't load. Make sure both shared models are present in the directory above and the file is a valid openWakeWord .onnx (input [1,16,96]).