Monitoring a user of a head-wearable electronic device
Systems, methods, and computer-readable media for monitoring a user of a head-wearable electronic device with multiple light-sensing assemblies.
1. A method of detecting head gestures at an electronic device including a plurality of light-sensing components being worn on a head of a user, the method comprising:
during a first training period in which the user performs a first head gesture,
collecting first training data from the plurality of light-sensing components,
calculating first training signal characteristics based on the first training data,
assigning at least one of the first training signal characteristics to either a first cluster of signal characteristics or a second cluster of signal characteristics, and
in accordance with a determination that a greater number of the first training signal characteristics have been assigned to the first cluster than to the second cluster, assigning the first cluster to the first head gesture; and
during a first gesture recognition period,
collecting first sensor data from the plurality of light-sensing components,
calculating first sensor signal characteristics based on the first sensor data, and
in accordance with a determination that more of the first sensor signal characteristics belong to the first cluster than to the second cluster, determining that the user has performed the first head gesture.
2. The method of claim 1 , further comprising:
during a second training period in which the user performs a second head gesture,
collecting second training data from the plurality of light-sensing components,
calculating second training signal characteristics based on the second training data,
assigning at least one of the second training signal characteristics to either the second cluster of signal characteristics or the first cluster of signal characteristics, and
in accordance with a determination that a greater number of the second training signal characteristics have been assigned to the second cluster than to the first cluster, assigning the second cluster to the second head gesture; and
during the first gesture recognition period,
in accordance with a determination that more of the first sensor signal characteristics belong to the second cluster than to the first cluster, determining that the user has performed the second head gesture.
3. The method of claim 2 , further comprising:
calculating first mean signal characteristics for the first cluster;
generating a first template for the first head gesture based on the first mean signal characteristics for the first cluster;
calculating second mean signal characteristics for the second cluster;
generating a second template for the second head gesture based on the second mean signal characteristics for the second cluster; and
comparing the first sensor signal characteristics to the first and second templates.
4. The method of claim 3 , wherein comparing the first sensor signal characteristics to the first and second templates comprises:
calculating a first distance from the first sensor signal characteristics to the first template and calculating a second distance from the first sensor signal characteristics to the second template;
in accordance with a determination that the first distance is shorter than the second distance, determining that the first sensor signal characteristics belong to the first cluster; and
in accordance with a determination that the second distance is shorter than the first distance, determining that the first sensor signal characteristics belong to the second cluster.
5. The method of claim 2 , further comprising:
during a third training period in which the user does not perform the first or second head gesture,
collecting third training data from the plurality of light-sensing components;
calculating third training signal characteristics based on the third training data;
assigning at least one of the third training signal characteristics to either a third cluster of signal characteristics, the second cluster of signal characteristics, or the first cluster of signal characteristics, and
in accordance with a determination that a greater number of the third training signal characteristics have been assigned to the third cluster than to the second cluster or the first cluster, assigning the third cluster to a third gesture different from the first or second head gestures; and
during the first gesture recognition period,
in accordance with a determination that more of the first sensor signal characteristics belong to the third cluster than to the first or second clusters, determining that the user has performed the third gesture.
6. The method of claim 5 , wherein the assigning of the first and second training signal characteristics to the first and second clusters of signal characteristics is performed using a k-means clustering algorithm.
7. The method of claim 6 , wherein the k-means clustering algorithm is also applied to the first sensor signal characteristics, and wherein determining whether the first sensor signal characteristics belong to the first cluster, the second cluster, or the third cluster is based on the k-means clustering algorithm.
8. The method of claim 1 , wherein calculating the first training signal characteristics comprises calculating at least one of the following:
an amplitude difference between a peak and a trough of the first training data;
a time difference between a peak and a trough of the first training data;
a maximum amplitude of the first training data;
a period between peaks of the first training data; and
a phase of the first training data.
9. The method of claim 1 , further comprising filtering heart rate frequencies from the first training data before calculating the first training signal characteristics based on the first training data.
10. A head-wearable electronic device, comprising:
a head-wearable housing structure;
a light-sensing assembly; and
a processor that is configured to detect head gestures performed by a user's head when wearing the head-wearable housing structure by,
during a first training period in which the user performs a first head gesture,
collecting first training data from a plurality of light-sensing components,
calculating first training signal characteristics based on the first training data,
assigning at least one of the first training signal characteristics to either a first cluster of signal characteristics or a second cluster of signal characteristics, and
in accordance with a determination that a greater number of the first training signal characteristics have been assigned to the first cluster than to the second cluster, assigning the first cluster to the first head gesture; and
during a first gesture recognition period,
collecting first sensor data from the plurality of light-sensing components,
calculating first sensor signal characteristics based on the first sensor data, and
in accordance with a determination that more of the first sensor signal characteristics belong to the first cluster than to the second cluster, determining that the user has performed the first head gesture.
11. The head-wearable electronic device of claim 10 , wherein the processor is further configured to detect the head gestures by:
during a second training period in which the user performs a second head gesture,
collecting second training data from the plurality of light-sensing components,
calculating second training signal characteristics based on the second training data,
assigning at least one of the second training signal characteristics to either the second cluster of signal characteristics or the first cluster of signal characteristics, and
in accordance with a determination that a greater number of the second training signal characteristics have been assigned to the second cluster than to the first cluster, assigning the second cluster to the second head gesture; and
during the first gesture recognition period,
in accordance with a determination that more of the first sensor signal characteristics belong to the second cluster than to the first cluster, determining that the user has performed the second head gesture.
12. The head-wearable electronic device of claim 11 , wherein the processor is further configured to detect the head gestures by:
calculating first mean signal characteristics for the first cluster;
generating a first template for the first head gesture based on the first mean signal characteristics for the first cluster;
calculating second mean signal characteristics for the second cluster;
generating a second template for the second head gesture based on the second mean signal characteristics for the second cluster; and
comparing the first sensor signal characteristics to the first and second templates.
13. The head-wearable electronic device of claim 12 , wherein the processor is configured to compare the first sensor signal characteristics to the first and second templates by:
calculating a first distance from the first sensor signal characteristics to the first template and calculating a second distance from the first sensor signal characteristics to the second template;
in accordance with a determination that the first distance is shorter than the second distance, determining that the first sensor signal characteristics belong to the first cluster; and
in accordance with a determination that the second distance is shorter than the first distance, determining that the first sensor signal characteristics belong to the second cluster.
14. The head-wearable electronic device of claim 11 , wherein the processor is further configured to detect the head gestures by:
during a third training period in which the user does not perform the first or second head gesture,
collecting third training data from the plurality of light-sensing components;
calculating third training signal characteristics based on the third training data;
assigning at least one of the third training signal characteristics to either a third cluster of signal characteristics, the second cluster of signal characteristics, or the first cluster of signal characteristics, and
in accordance with a determination that a greater number of the third training signal characteristics have been assigned to the third cluster than to the second cluster or the first cluster, assigning the third cluster to a third gesture different from the first or second head gestures; and
during the first gesture recognition period,
in accordance with a determination that more of the first sensor signal characteristics belong to the third cluster than to the first or second clusters, determining that the user has performed the third gesture.
15. The head-wearable electronic device of claim 14 , wherein the processor is configured to assign the first and second training signal characteristics to the first and second clusters using a k-means clustering algorithm.
16. The head-wearable electronic device of claim 15 , wherein the k-means clustering algorithm is also applied to the first sensor signal characteristics, and wherein determining whether the first sensor signal characteristics belong to the first cluster, the second cluster, or the third cluster is based on the k-means clustering algorithm.
17. The head-wearable electronic device of claim 10 , wherein the processor is configured to calculate the first training signal characteristics by calculating at least one of the following:
an amplitude difference between a peak and a trough of the first training data;
a time difference between a peak and a trough of the first training data;
a maximum amplitude of the first training data;
a period between peaks of the first training data; and
a phase of the first training data.
18. The head-wearable electronic device of claim 10 , wherein the processor is further configured to detect the head gestures by filtering heart rate frequencies from the first training data before calculating the first training signal characteristics based on the first training data.