Method and apparatus for real-time gunshot detection and reporting

1. A method comprising:

detecting, in an audio signal received by an audio receiver comprising a plurality of audio receiving elements, at least one gunshot;

determining, based on the plurality of audio receiving elements, an audio location of the at least one gunshot;

identifying, based on the audio location and on information that identifies locations and orientations of a plurality of cameras, at least two cameras that can capture an image of the audio location;

sending control signals to the at least two cameras to cause the at least two cameras to capture at least two corresponding images of the audio location;

based on each of the at least two corresponding images, identifying a gunshot location; and

transmitting a gunshot location identifier that identifies the gunshot location to a destination,

wherein identifying the gunshot location further includes:

determining a global positioning system (GPS) location of each of the at least two cameras;

generating a depth map of the audio location depicted in each of the at least two corresponding images; and

identifying the gunshot location based on the depth map and the GPS location of each of the at least two cameras.

2. The method of claim 1 , wherein the gunshot location identifier comprises GPS coordinates.

3. The method of claim 1 , further comprising:

detecting, in at least one of the two corresponding images, a gun; and

determining a location of the gun based on the depth map; and

wherein the gunshot location comprises the location of the gun.

4. The method of claim 1 , further comprising:

detecting, in at least one of the two corresponding images, a gun;

extracting, from the at least one of the two corresponding images, a gun image that depicts the gun; and

transmitting, to the destination, the gun image.

5. The method of claim 1 wherein the detecting the at least one gunshot further comprises:

inputting the audio signal into a gunshot detection machine-learned model that has been trained using a plurality of different urban sounds and a plurality of different gunshot sounds corresponding to a plurality of different guns; and

receiving, from the gunshot detection machine-learned model, an indication that the audio signal comprises a gunshot.

6. The method of claim 1 wherein detecting the at least one gunshot further comprises:

detecting, based on the audio signal, a plurality of gunshots;

determining that the plurality of gunshots has occurred within a predetermined period of time; and

wherein determining the audio location of the at least one gunshot comprises, in response to determining that the plurality of gunshots has occurred within the predetermined period of time, determining the audio location of the plurality of gunshots.

7. The method of claim 1 further comprising:

analyzing the audio signal to identify a type of gun associated with the at least one gunshot; and

transmitting, to the destination, a gun type identifier that identifies the type of gun.

8. The method of claim 7 , wherein analyzing the audio signal to identify the type of gun further comprises:

inputting the audio signal into a gun type machine-learned model that has been trained using a plurality of different urban sounds, a plurality of different gunshot sounds corresponding to a plurality of different guns, and a plurality of gun type identifiers that identify the plurality of different guns; and

receiving, from the gun type machine-learned model, an output comprising the gun type identifier.

9. The method of claim 7 further comprising:

detecting, based on the audio signal, a plurality of gunshots including the at least one gunshot;

determining, based on the plurality of gunshots, a rate of fire; and

transmitting, to the destination, rate of fire information that identifies the rate of fire.

10. The method of claim 1 further comprising:

identifying, in at least one of the at least two corresponding images, a gun;

identifying, in the at least one of the at least two corresponding images, one or more faces that are within a predetermined distance of the gun;

extracting from the at least one of the at least two corresponding images, for each face of the one or more faces, a corresponding one or more face images that depict the face; and

transmitting, to the destination, the one or more face images.

11. The method of claim 1 further comprising:

capturing, from at least one camera of the at least two cameras, an image of a setting of the audio location in a scene corresponding to the at least one camera of the at least two cameras; and

transmitting, to the destination, the image.

12. The method of claim 1 further comprising:

capturing, from at least one camera of the at least two cameras, a video recording of a setting of the audio location in a scene corresponding to the at least one camera of the at least two cameras; and

transmitting, to the destination, the video recording or a URI to the video recording.

13. A gunshot detection system, comprising:

a communications interface; and

a processor device coupled to the communications interface and configured to:

detect, in an audio signal received by an audio receiver comprising a plurality of audio receiving elements, at least one gunshot;

determine, based on the plurality of audio receiving elements, an audio location of the at least one gunshot;

identify, based on the audio location and on information that identifies locations and orientations of a plurality of cameras, at least two cameras that can capture an image of the audio location;

send control signals to the at least two cameras to cause the at least two cameras to capture at least two corresponding images of the audio location;

based on each of the at least two corresponding images, identify a gunshot location; and

transmit a gunshot location identifier that identifies the gunshot location to a destination,

wherein identifying the gunshot location further includes:

determining a location of each of the at least two cameras;

generating a depth map of the audio location depicted in each of the at least two corresponding images; and

identifying the gunshot location based on the depth map and the location of each of the at least two cameras.

14. The gunshot detection system of claim 13 , wherein the processor device is further configured to:

detect, in at least one of the two corresponding images, a gun; and

determine a location of the gun based on the depth map; and

wherein the gunshot location comprises the location of the gun.

15. The gunshot detection system of claim 13 , wherein the processor device is further configured to:

analyze the audio signal generated by the audio receiver to identify a type of a gun associated with the at least one gunshot; and

transmit, to the destination, a gun type identifier that identifies the type of the gun.

16. The gunshot detection system of claim 13 , wherein the processor device is further configured to:

identify, in at least one of the at least two corresponding images, a gun;

identify, in the at least one of the at least two corresponding images, one or more faces that are within a predetermined distance of the gun;

extract from the at least one of the at least two corresponding images, for each face of the one or more faces, a corresponding one or more face images that depicts the face; and

transmit, to the destination, the one or more face images.