Event-detection confirmation by voice user interface
Systems and methods for event-detection confirmation are disclosed. For example, a sensing device may generate sensor data, such as from radar, and the sensor data may be utilized to determine if a predefined event as occurred. Event-confirmation operations may then be performed, such as by utilizing acoustic-event detection techniques and/or natural language understanding techniques. When occurrence of an event is confirmed, such as to a certain confidence level, one or more actions may be taken, such as sending a notification to another device and/or establishing a communication channel with another device, such as a device associated with emergency services, family members, friends, and/or neighbors.
1 . A method comprising:
receiving, at a first device and from a second device disposed in an environment of the first device, first sensor data;
receiving, at the first device and from a third device disposed in the environment, second sensor data, wherein the first sensor data is associated with a first sensor and the second sensor data is associated with a second sensor that differs from the first sensor;
generating input data from the first sensor data and the second sensor data, the input data configured to be utilized by an on-device model of the first device to detect a fall event in the environment;
querying the on-device model, utilizing the input data, for a determination of whether the fall event has occurred; and
selecting at least one of the second device or the third device to perform an action associated with detection of the fall event by the on-device model of the first device.
2 . The method of claim 1 , wherein:
the first device, the second device, and the third device are co-located in the environment;
the second device is a wearable device; and
the third device is a communal device.
3 . The method of claim 1 , wherein:
the first sensor data comprises data indicating that the second device has moved in a manner corresponding to the fall event; and
the second sensor data comprises audio data received from the environment indicating an acoustic event associated with a fall.
4 . The method of claim 1 , wherein the on-device model comprises a machine learning model configured to receive the input data in a form of a query and to generate output data indicating the determination of whether the fall event has occurred.
5 . The method of claim 1 , wherein:
the first sensor data comprises data indicating that the second device has moved in a manner corresponding to the fall event; and
the second sensor data comprises image data received from the environment indicating depiction of a person that has fallen.
6 . The method of claim 1 , wherein:
the first sensor data comprises data indicating that the second device has moved in a manner corresponding to the fall event; and
the second sensor data comprises radar data received from the environment indicating the fall event.
7 . The method of claim 1 , wherein the action includes outputting a notification, and the method further comprises causing the notification to be output by causing a component of the second device to vibrate.
8 . The method of claim 1 , wherein the action includes automatically establishing a communication channel between the second device and an emergency service.
9 . The method of claim 1 , wherein the action includes outputting a notification by causing display of a user interface on the second device requesting confirmation that the fall event occurred.
10 . The method of claim 1 , wherein the action includes outputting a notification by causing output of audio on the third device requesting speech input confirming that the fall event occurred.
11 . A first device comprising:
one or more processors; and
non-transitory computer-readable media storing instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising:
receiving, at the first device and from a second device disposed in an environment of the first device, first sensor data;
receiving, at the first device and from a third device disposed in the environment, second sensor data, wherein the first sensor data is associated with a first sensor and the second sensor data is associated with a second sensor that differs from the first sensor;
generating input data from the first sensor data and the second sensor data, the input data configured to be utilized by an on-device model of the first device to detect a fall event in the environment;
querying the on-device model, utilizing the input data, for a determination of whether the fall event has occurred; and
selecting at least one of the second device or the third device to perform an action associated with detection of the fall event by the on-device model of the first device.
12 . The first device of claim 11 , wherein:
the first device, the second device, and the third device are co-located in the environment;
the second device is a wearable device; and
the third device is a communal device.
13 . The first device of claim 11 , wherein:
the first sensor data comprises data indicating that the second device has moved in a manner corresponding to the fall event; and
the second sensor data comprises audio data received from the environment indicating an acoustic event associated with a fall.
14 . The first device of claim 11 , wherein the on-device model comprises a machine learning model configured to receive the input data in a form of a query and to generate output data indicating the determination of whether the fall event has occurred.
15 . The first device of claim 11 , wherein:
the first sensor data comprises data indicating that the second device has moved in a manner corresponding to the fall event; and
the second sensor data comprises image data received from the environment indicating depiction of a person that has fallen.
16 . The first device of claim 11 , wherein:
the first sensor data comprises data indicating that the second device has moved in a manner corresponding to the fall event; and
the second sensor data comprises radar data received from the environment indicating the fall event.
17 . The first device of claim 11 , wherein the action includes outputting a notification, and the operations further comprise causing the notification to be output by causing a component of the second device to vibrate.
18 . The first device of claim 11 , wherein the action includes automatically establishing a communication channel between the second device and an emergency service.
19 . The first device of claim 11 , wherein the action includes outputting a notification by causing display of a user interface on the second device requesting confirmation that the fall event occurred.
20 . The first device of claim 11 , wherein the action includes outputting a notification by causing output of audio on the third device requesting speech input confirming that the fall event occurred.