A lightweight, simple-to-use, wake word listener, including all the tools to train your own custom wake word using a recurrent neural network.
Mycroft Precise monitors an audio stream (usually a microphone) and when it recognizes a specific phrase it triggers an event. For example, by default users of the Mycroft Voice Assistant are using a Precise model trained to spot the phrase "Hey Mycroft". When Precise recognizes this phrase it puts the Mycroft Voice Assistant into command mode performing speech recognition on whatever is next said by the person using the device.
Mycroft Precise is fully open source and can be trained to recognize any short-phrase or sound, from a name to a cough.
The default "Hey Mycroft" model is included. To try it out, run:
mycroft-precise.listen /snap/mycroft-precise/current/hey-mycroft/hey-mycroft.pb
USAGE
Running the listener
mycroft-precise - Alias for mycroft-precise.enginemycroft-precise.engine - Run a model on raw audio data from stdinmycroft-precise.listen - Run a model on microphone audio inputmycroft-precise.listen-pocketsphinx - Run the PocketSphinx listenerData collection
mycroft-precise.collect - Record audio samples for use with Precisemycroft-precise.add-noise - Create a duplicate dataset with added noiseTraining
mycroft-precise.train - Train a new model on a datasetmycroft-precise.train-generated - Train a model on infinitely generated batchesmycroft-precise.train-incremental - Train a model to inhibit activation by marking false activations and retrainingmycroft-precise.train-optimize - Use black box optimization to tune model hyperparametersmycroft-precise.train-sampled - Train a model, sampling data points with the highest loss from a larger datasetEvaluation and analysis
mycroft-precise.test - Test a model against a datasetmycroft-precise.test-pocketsphinx - Test PocketSphinx against a datasetmycroft-precise.eval - Evaluate a list of models on a datasetmycroft-precise.calc-threshold - Update the threshold values of a model for a dataset to make the sensitivity more accurate and linearmycroft-precise.graph - Show ROC curves for a series of modelsmycroft-precise.simulate - Simulate listening to long chunks of audio to find unbiased false positive metricsModel conversion
mycroft-precise.convert - Convert wake-word model from Keras to TensorFlowYou are about to open
Do you wish to proceed?
Thank you for your report. Information you provided will help us investigate further.
There was an error while sending your report. Please try again later.
Snaps are applications packaged with all their dependencies to run on all popular Linux distributions from a single build. They update automatically and roll back gracefully.
Snaps are discoverable and installable from the Snap Store, an app store with an audience of millions.
Snap can be installed from the command line on openSUSE Leap 15.x and Tumbleweed.
You need first add the snappy repository from the terminal. Choose the appropriate command depending on your installed openSUSE flavor.
Tumbleweed:
sudo zypper addrepo --refresh https://download.opensuse.org/repositories/system:/snappy/openSUSE_Tumbleweed snappy
Leap 15.x:
sudo zypper addrepo --refresh https://download.opensuse.org/repositories/system:/snappy/openSUSE_Leap_15.6 snappy
If needed, Swap out openSUSE_Leap_15. for, openSUSE_Leap_16.0 if you’re using a different version of openSUSE.
With the repository added, import its GPG key:
sudo zypper --gpg-auto-import-keys refresh
Finally, upgrade the package cache to include the new snappy repository:
sudo zypper dup --from snappy
Snap can now be installed with the following:
sudo zypper install snapd
You then need to either reboot, logout/login or source /etc/profile to have /snap/bin added to PATH.
Additionally, enable and start both the snapd and the snapd.apparmor services with the following commands:
sudo systemctl enable --now snapd
sudo systemctl enable --now snapd.apparmor
To install Mycroft Precise, simply use the following command:
sudo snap install mycroft-precise --edge
Browse and find snaps from the convenience of your desktop using the snap store snap.
Interested to find out more about snaps? Want to publish your own application? Visit snapcraft.io now.