Video analytics alone do not yield a high rate of incident detection. Even though cameras now can detect a range of bad behaviors, the video is mainly used to review interactions after the event has occurred. This is where audio analytics come into the picture.
Surveillance technology has grown to play an important role in delivering critical details to investigators and first responders. Many security personnel rely on CCTV and video analytics to learn important information about crime scenes. However, since video surveillance is primarily reactive, audio can be used to proactively deter altercations.
By integrating audio analytics with an existing security solution, such as one of Milestone System’s VMS solutions, a manager can assess the verbal tone of the situation, and in cases where there are hostile voices, take measures to prevent the conflict from escalating into a fight.
How do audio analytics work?
Audio analytics provide the ability to analyze and identify a sound. Similar to how the human ear processes audio, the analytic software analyzes sounds through advanced algorithms and classifies it into a predetermined category such as aggression or breaking glass. Imagine the applications for fast alerts in retail robberies, gang-related tensions, crowded events, and patients in healthcare institutions.
For security systems, audio analytics add accuracy and offer the following benefits:
- Proactive Response: Audio analytics identify verbal aggression, giving valuable time to security personnel for speedy intervention before it turns into a violent outbreak. Analytics ensure that no incident goes unnoticed, and by the same token, that all incidents are analyzed to avoid false alarms.
- Cost-Effective Investment: By using the technology, more incidents are prevented, as well as their costly consequences. Installing audio analytics software is far less expensive than investigating an assault, interviewing employees, paying court fees, etc.
- Monitoring Efficiency: Usually end users have to hire multiple central station guards to monitor several video screens. When there are audio and video analytics, less personnel are required because not all cameras need to be monitored simultaneously. Instead, staff can be directed to look at the relevant camera as a follow-up from an audio detection and alert.
- Omnidirectional Capability: Using omnidirectional microphones, audio analytics analyze sound from 360 degrees whereas video alone usually captures information from a single angle.
- Additional Evidence: Audio events provide additional, unique information concerning specific incidents. An analytics program makes it easy to listen back to the sound that triggered an alarm. (For detection, the audio does not need to be recorded, but it can be recorded if so desired by the end user.)
- Real-time Response: The system alerts personnel at the precise moment the audio event occurs.
- Confidentiality and Robustness: Language content is not analyzed during sound detection and thus enhances privacy. In other words, the software does not listen to words, but distortion and physiological changes in the voice. This makes the technology fully international and not language dependent.
- Sensitivity and Flexibility: In different environments, different levels of aggression can be tolerated. In a sport stadium, an exciting match will stir the spectators – a queue at the Department of Motor Vehicles is an entirely different setting. When it is crucial to not miss an occurrence, the sensitivity can be set to a higher level. In places where unwanted alarms are more likely to occur, the sensitivity level can be reduced.