Skip to content
OmnilertOct 2, 2023 11:35:38 PM8 min read

AI Visual Gun Detection Takes Center Stage at GSX 2023

GSX 2023 is a wrap and AI visual gun detection technology continued to be one of the hottest topics at the show. Just check out TheTechieGuy’s coverage in this YouTube video titled “This tech can help stop attacks - but why isn't it EVERYWHERE?

The host, Liron Segev, was on a mission to find which companies at the show were using technology to keep us safe in the real-world... and he instantly zoned in on Omnilert. Liron called Omnilert Gun Detect one of the coolest technologies he has seen, touting it as an “amazing system for early detection of someone holding a weapon.” He then went on to demo the system himself in the Omnilert booth, where he saw how quickly and accurately the system could detect a brandished weapon.  As he pointed out, most shooters are first visible with their weapon when they are outside a facility and if they can be detected out in a parking lot for example, there will be a much better chance of securing the facility before they enter the building. 

As Liron stated “why is this technology not in every single school and in every single work place?” To see Liron demo the product in real-time and learn more about how it can keep people safe from active shooters, you can watch the full video below:


More Show Highlights

GSX 2023 was one of our busiest shows yet, with interest coming from nearly every sector of the industry, including government, system Integrators, K-12 & higher education, banking, manufacturing, military and more.  While we can’t cover all the exciting developments around our technology from the show, below are a few of the more significant ones:

Omnilert was the only visual gun detection solution providing a live demo of its software
With 13,000 people walking the show floor, what a great way to highlight our high accuracy rate while running continual, live demonstrations of our AI visual gun detection software. They key to achieving high accuracy comes from the way AI models are trained.  Unlike other solutions, we train our models using 100% real video footage from actual live settings, both indoors and outdoors from sources around the world.  This means better results versus training models with green screens or using synthetic data.  Maybe that is why we were the only vendor to actually show a live demo.

Evolv Partners with Omnilert
Heading into the show, Evolv announced that it had integrated Omnilert’s AI visual gun detection software into its newly announced Evolv Extend.  Leveraging Omnilert’s open platform and flexible monitoring capabilities that allow Evolv to choose where to send detection alerts, Evolv Extend adds yet another layer of protection empowering onsite security operators to take immediate action to help prevent potentially catastrophic active shooter incidents.  This partnership was well received at the GSX show, with consensus that this move combined the leading AI technologies for concealed weapon and brandished gun detection.  Thanks to Evolv for the great demos they were showcasing in their booth.

Omnilert’s Director of AI takes the Stage
Chad Green, our Director of AI, gave a powerful talk on the importance of combining visual gun detection intelligence with a nearly instantaneous automated response and communication as the key factor to save lives in moments of gun violence or mass shootings. Below are highlights from his presentation, which was titled “Detection & Response Activation—Critical Components for Active Shooter Technology Solutions” as well as key points he provided during the non-stop Q& A session:

  • There is no greater threat today than Gun Violence – with the statistics speaking for themselves.
  • The first minute of any emergency situation is the most important and the action taken during those seconds determine the best outcomes.
  • The average duration of an active shooter incident is 8 minutes, yet the time it takes for police to arrive is also often 8 minutes because it generally takes a person 5 minutes to call 911.
  • Today, there are several available technologies for gun detection that can be layered together -  human detection, metal detection and visual detection.
    • The biggest difference between AI detection and human detection is that AI doesn’t fatigue or take breaks and can monitor as many camera feeds as needed, without distraction.
  • AI analytics is only as good as the data it was trained on. That is why it’s so important to make sure AI is trained on cameras in the real-world with random people and real backgrounds versus studios with models in front of green screens or realistic 3D animations such as those used in video games.
  • AI visual gun detection could have helped many prior active shooting events such as Parkland, Nashville and Sandy Hook. In every case, detection and verification happening instantly was key – but the response and notification that followed would have been just as important.
    • Real-time situational intelligence could have provided first responders with actionable details such as the location of the shooter, how many shooters there were, what type of weapon they had, etc. And when police arrive at campus, the Omnilert platform will also recognize first responders with weapons, providing real-time information on where police are in relation to the shooter.
  • AI can turn existing cameras into active weapons detectors. For large areas, it is important to have enough cameras to ensure all areas are being monitored.  It’s also important to make sure your cameras provide the right pixels per foot for clarity, such as 30-40 pixels per foot.  Cameras should also be evaluated not only for their performance during the day, but also at night in the dark.
  • With AI gun detection, every camera installed inside and outside a facility is being monitored in real-time in parallel. This is the same whether there are 10 cameras or 1,000.
  • AI visual gun detection technology should integrate into existing systems such as VMS solutions that can take the alert and track it through their own platform. That is why Omnilert has partnered with so many providers. Omnilert AI Gun Detect also provides the customer the flexibility to decide who they want to get the alerts once a weapon has been detected. This could be internal security teams, Omnilert’s monitoring service or a third-party monitoring company.
  • While we can’t prevent active shooter incidents from happening, organizations can follow these best practices today to be prepared in the future:
    • Establish crisis communications policies
    • Inventory safety systems for automation
    • Review best practices
    • Assemble your security and crisis response teams
    • Develop agreed-upon scenarios
    • Training and exercises
    • Establish decision points

To learn more about Omnilert Gun Detect, visit the Active Shooter Solution page. If you are interested in learning more about AI gun detection and our visual active shooter solution contact us now.