MV Audio Detection
All MV cameras coming after 1st generation are capable of processing powerful analytics on the camera itself and transmitting this metadata to the Meraki cloud. This revolutionary architecture dramatically reduces the cost and complexity of gathering detailed analytics in any environment.
With this architecture, all MV cameras have the ability to transmit motion metadata to the Meraki cloud by enabling Motion Search, a powerful tool in retrieving video and Motion Heatmaps, which is part of MV analytics.
On the second-generation cameras, one of the new and exciting capabilities of this platform is the ability to do machine learning-based analytics. With this comes audio detection. This allows you to understand the audio levels in your environments and detect for specific emergency-related events.
This feature allows users to use the microphone embedded in MV cameras (except for 1st generation) as a sensor for their environment. The audio detection feature provides two key capabilities:
- detection of two specific audio classes: fire alarms and emergency sirens
- detection of ambient noise levels, as measured in decibels (dB)
Detection of these audio classes provides situational awareness to potential emergency events, and understanding ambient noise levels can help to monitor for unusual events or disturbances in offices, workspaces, factory floors, and retail environments.
This feature is presently in Beta and may change over time as we work to improve!
MV firmware version 4.12+ required.
In order to configure audio detection on your MV camera, navigate to the Settings > MV Sense tab, ensure MV Sense is applied to this camera, and enable the feature. Be certain to save your configuration changes. Once enabled, the camera will immediately begin processing the audio stream.
In order to receive and monitor the output of MV audio detection, you can subscribe to the appropriate MQTT topics. See the MV Sense API section for more detailed information.
Privacy by Design, Privacy by Default
Audio detection is performed independent from audio recording. Enabling or disabling audio detection will not affect any audio recording policies on any cameras. See this article to learn more about audio recording on 2nd generation MV cameras.
Audio detection does not store the input audio signal in any way. You must enable audio detection under a MV Sense license before the camera will begin detecting from its input audio stream.
MV Sense API
The final way to interact with MV audio detection analytics is to use the API endpoints provided with MV Sense to build intelligent business solutions. Read the MV Sense article for more information.
How does it work?
Software on the camera processes the input audio stream, using windowed sub-sampling. These samples are then converted into visual representations of the input audio signal and processed by a model trained for specific audio classes. These audio detections can be streamed for detailed analysis and storage via MQTT. This functionality requires a MV Sense license to be applied to the camera. Our audio detection is driven by computer vision and machine learning.
What do you mean by "machine learning"?
Meraki smart cameras use deep learning, a type of machine learning at the forefront of artificial intelligence research, to drive our computer vision object detection. The smart camera development teams continually show a computer thousands of examples of what objects look like and it "learns" how to identify them more and more accurately over time. The model improves as we provide it with additional training data.
The smart camera analytics models are trained on data that is legally owned or licensed by Cisco Meraki, and only uses data from customers who have explicitly opted-in to continually improve our models.
Does this take any extra bandwidth? Do I need another server?
All of the audio processing is done right on the camera's processor. Like all Meraki products, the hardware sends small amounts of metadata back to the dashboard for further processing and storage. The MQTT messaging data may be sent to any MQTT broker resource of your choosing.
Common Deployment Challenges
When choosing how to deploy your MV for audio detection, consider the following challenges:
Avoid deploying the camera in locations surrounded with highly reflective acoustic surfaces (like concrete, metals, glass).
Certain anomalous audio signals (bird calls, warning/notification noises, etc) may be momentarily detected as a fire alarm or emergency siren.
There are no known deployment challenges to the decibel level output from a MV camera, provided it is installed in a location where the microphone will have as few obstructions to the environment as possible.
If you observe repeatable issues with audio detection or analytics, please make a wish to let us know how we can improve!