| CPC H04S 7/303 (2013.01) [G02B 27/0176 (2013.01); G06N 3/08 (2013.01); H04R 5/04 (2013.01); H04S 7/40 (2013.01); G02B 2027/0138 (2013.01); G02B 2027/0178 (2013.01); H04S 2400/11 (2013.01); H04S 2420/01 (2013.01)] | 20 Claims |
|
1. An eyewear device comprising:
a microphone system;
a presentation system;
a support structure configured to be head-mounted on a user, the support structure supporting the microphone system and the presentation system; and
a processor, a memory, and programming in the memory, wherein execution of the programming by the processor configures the eyewear device to:
capture, with the microphone system, audio information of an environment surrounding the eyewear device;
identify an audio signal within the audio information, wherein to identify the audio signal within the audio information the processor applies a signal discrimination filter to the audio information;
detect a direction of the audio signal with respect to the eyewear device, wherein to detect the direction of the audio signal with respect to the eyewear device the processor applies a beam forming algorithm;
classify the audio signal into one of a plurality of predefined classifications, each of the plurality of predefined classifications associated with a respective application for presentation by the presentation system, wherein to classify the audio signal into one of a plurality of predefined classifications the processor applies a trained convolutional neural network (CNN) to the audio signal;
monitor a direction processing timestamp corresponding to detecting the direction of the audio signal;
monitor a CNN processing timestamp corresponding to applying the trained CNN to the audio signal;
correlate the direction processing timestamp and the CNN processing timestamp; and
present, by the presentation system, the respective application associated with the one of the plurality of predefined classifications responsive to the direction of the audio signal, wherein presenting the respective application associated with the one of the plurality of predefined classifications is further responsive to the correlated CNN processing timestamp and direction processing timestamp.
|