Gurgaon-based Video AI implementation enabler Staqu Technologies has recently launched a new audio analytics feature for Jarvis, its video AI engine. Aimed at addressing the informational blind-spots that currently exist in the CCTV monitoring and security space, the upgrade enables Jarvis to evaluate audio inputs from CCTV cameras, in addition to the video feed, to accurately identify and classify potentially dangerous events.
The fundamental idea behind Staqu’s latest feature rollout is this: CCTV cameras can capture the ambient audio data from outside their direct field of vision that can be used to identify any disturbance or incongruous event. For instance, the sound of a glass shattering outside a physical store can denote burglaries, acts of vandalism, or sabotage in progress, while gunshots, screaming and loud shouting typically indicate physical violence and a clear and immediate danger to life.
As per Staqu, Jarvis’s audio analytics feature can evaluate this input to conduct accurate ‘scene or event classification’ and notify human responders in real-time to trigger the most appropriate situational response. The company claims that it will improve the operational efficacy of law enforcement agencies and private security teams, enabling them to drastically shorten their response time to such events and contain them before they can escalate.
With the rollout, Jarvis will also provide unique audio-based search functionality on its platform; one can search for any video activity just by presenting an audio query to Jarvis. According to Staqu, this can help business owners and other authorised users – such as administrators and HR professionals – identify individuals not following standard operating procedures such as COVID-19 protocols, safety protocols, security guidelines, etc. at the workplace.