IREXDocs
EN

AudioTrack

AudioTrack detects gunshots, screaming, glass breaking, and loud noises using the camera microphone.

Meet AudioTrack

Audio Recording Requirements

Parameter Requirement
Supported audio codecs AAC, G.711 μ-law/A-law, G.726, PCM
Bandwidth 80 Hz to 16 kHz
Bit depth 8 to 64 bits

Computational Complexity

The algorithm is rated 1 on a 5-point complexity scale. Average CPU usage is about 20%, and the load can vary with the quality of the audio channel and the processor.

Set Up Distinctive Sound Detection

  1. Switch on the audio event types to generate under the EVENTS (Loud sound, Gunshot, Glass breaking, Scream).
    2. AudioTrack detector settings with the EVENTS list showing Loud sound, Gunshot, Breaking glass, and Scream all enabled at Normal priority
  2. For each audio event type, select a priority rating value.
  3. Drag the Audio sensitivity slider to adjust sensitivity to the event sound volume relative to the background noise level.
  4. Click Apply.

Any loud sound the module failed to identify as one of Gunshot/Glass breaking/Scream is reported simply as "Loud sound".