A system and method are disclosed that may predict the intent of a volume up/down event in a phone device, and direct the action so that the intended setting is adjusted. The method includes two steps. In the first step, a machine learning model deployed on the user’s device is trained to predict the probability that media sound is about to be played from the device. In the second step, when the volume button is pressed, the machine learning model, depending on the predicted probability that media sound is about to be played, may delegate the action to either the media volume setting or to the ringtone volume setting. Further, the user may additionally personalize the model. The advantage of the system includes smart de-multiplexing of multiple settings through one input endpoint control. The system requires no additional system UI elements.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.