App of the AllSpeak project. Is an hybrid App developed with Ionic v1, mailny relying on the cordova-plugin-speechrecognition plugin.
It interacts with the https://api.allspeak.eu server to access its tensorflow training routines. User must be previously enabled (receiving an individual code) to access these features.
The App manages the following aspect:
- register device configuration on the server
- manage several vocabularies (list of commands to be recognized)
- prepare data for training
- records user voice while repeating the commands, both for playbacking them when are recognized and to train the neural net
- use smartphone microphone or a bluetooth headset