Automatic sensor orientation calibration
A wearable audio device is provided. The wearable audio device includes a sensor, such as an IMU and a controller. The sensor is configured to capture rotational motion data. At least a portion of the captured rotational motion data corresponds to head motion of a user. The sensor is further configured to generate a sensor orientation of the sensor based on the rotational motion data. The controller is configured to (1) receive the rotational motion data and the sensor orientation from the sensor; (2) generate, based on the rotational motion data, an orientation calibration parameter; and (3) map the sensor orientation to a head orientation of the user based on the orientation calibration parameter.
1 . A wearable audio device, comprising:
a sensor configured to:
capture rotational motion data, wherein at least a portion of the captured rotational motion data corresponds to head motion of a user; and
generate a sensor orientation of the sensor based on the rotational motion data; and
a controller configured to:
receive the rotational motion data and the sensor orientation from the sensor;
generate, based on the rotational motion data, an orientation calibration parameter; and
map the sensor orientation to a head orientation of the user based on the orientation calibration parameter, wherein the controller is further configured to:
calculate, based on the rotational motion data, a series of rotation axes, wherein each of the series of rotation axes corresponds to one of a series of event periods during a movement period;
determine a rotational dispersion of the series of rotation axes; and
determine the orientation calibration parameter based on the rotational motion data if the rotational dispersion is within a dispersion threshold.
2 . The wearable audio device of claim 1 , wherein the sensor is an inertial measurement unit (IMU).
3 . The wearable audio device of claim 1 , wherein the rotational motion data comprises angular velocity.
4 . The wearable audio device of claim 1 , wherein the head motion comprises a yaw motion.
5 . The wearable audio device of claim 1 , wherein the head motion comprises a pitch rotation.
6 . The wearable audio device of claim 1 , wherein the dispersion threshold is less than or equal to 10 degrees.
7 . The wearable audio device of claim 1 , wherein the dispersion threshold is determined by a neural network model trained by historic rotation data.
8 . The wearable audio device of claim 1 , wherein the movement period is less than one minute.
9 . The wearable audio device of claim 1 , wherein the sensor orientation is defined by a sensor x-axis, a sensor y-axis, and a sensor z-axis.
10 . The wearable audio device of claim 1 , wherein the wearable audio device is an earbud.
11 . A method for automatically calibrating a sensor orientation of a sensor of a wearable audio device, comprising:
capturing, via the sensor, rotational motion data, wherein at least a portion of the captured rotational motion data corresponds to head motion of a user;
generating, via the sensor, the sensor orientation of the sensor based on the rotational motion data;
generating, based on the rotational motion data, an orientation calibration parameter; and mapping the sensor orientation to a head orientation of the user based on the orientation calibration parameter, wherein calibrating the sensor orientation of the sensor further comprises:
calculating, based on the rotational motion data, a series of rotation axes, wherein each of the series of rotation axes corresponds to one of a series of event periods during a movement period;
determining a rotational dispersion of the series of rotation axes; and
determining the orientation calibration parameter based on the rotational motion data if the rotational dispersion is within a dispersion threshold.
12 . The method of claim 11 , wherein the sensor is an inertial measurement unit (IMU).
13 . The method of claim 11 , wherein the rotational motion data comprises angular velocity.
14 . The method of claim 11 , wherein the head motion comprises a yaw motion.
15 . The method of claim 11 , wherein the head motion comprises a pitch motion.
16 . The method of claim 12 , wherein the dispersion threshold is less than or equal to 10 degrees.
17 . The method of claim 12 , wherein the dispersion threshold is determined by a neural network model trained by historic rotation data.
18 . The method of claim 11 , wherein the sensor orientation is defined by a sensor x-axis, a sensor y-axis, and a sensor z-axis.