Sound Events Detector: AI-based recognition of sounds
“Sound Events Detector” is a module in Xeoma based on Artificial Intelligence and deep machine learning. This module is designed to recognize 5 types of sound events: glass break, alarm, baby cry, screams and gunshots.
Combination with various reactions or detectors of Xeoma makes it a powerful automation tool both in private or corporate life.
Now your camera system can not only see things – it can hear events as well!
Detect intrusion before the burglar even appears in camera view – by the sound of shattering glass.
Your child has waken up and is crying? The system will not only notice that, but proactively notify you.
The alarm on your car has gone off in the parking lot? You don’t have to be hearing that. Xeoma will hear it for you – and let you know!
With Xeoma’s pre-record option you can also see video footage for seconds – or even minutes – before the event was detected!
In many boutiques, shopfront’s windows allow much easier and quicker access inside than doors. The same goes for offices, and private property. Quickly detect intrusion by the sound of breaking glass – even before the cameras see intruders!
Extremely useful for police in safe city programs. Have city cameras not just record 24/7 video footage but also recognize if there are shot fired in the streets. Pre-record feature will help unravel events before the weapon had been engaged.
Recognition of car alarm sound can be useful in city safety systems, in businesses (paid parking lots) and even in long-range private residences.
All range of screams, from all genders can be recognized in a Xeoma system, and a pre-set reaction will be triggered instantly. Perfect for city surveillance so that a victim of attack (or, the attacker) is heard even when out of plain sight.
Unlike regular baby monitors that only transmit video as is, Xeoma’s smart recognition of the sound of baby cry can send you a notification that you will actually notice as soon as the baby wakes up and needs your attention.
Using knowledge and experience gathered in data sets, the artificial intelligence of Xeoma’s sound recognition module is constantly “listening” to the camera stream for noises of a certain type, based on specific characteristics typical for that type.
It can be used in various situations, for example:
Police/safe city: city cameras are usually used for 24/7 record only which is good for post-incident investigation. But what if city surveillance can also be a proactive tool for real-time detection and reaction to unfolding criminal acts? The ‘Sound Events Detector’ can help quickly react to gunshots or screams of citizens, attacked or attacking.
Hint: use Xeoma’s Emotions Recognition to predict acts of aggression based on mood and behavior analysis.
Business (offices, warehouses, stores, etc.): regular intrusion systems are great, but now you don’t have to cover every square inch of the territory with cameras to not miss intrusion. If one of the few cameras hears the sound of breaking glass, even in off-hours, you will be notified. It’s perfect with PTZ cameras that cover a fast territory with their automatic tours, but can regularly miss something if the camera was looking in another direction at that moment.
Hint: with Xeoma’s “Move to PTZ Preset” module, an overview camera can be rotated automatically to focus on the source of the sound.
Banking/finance: instant panic button upon “hearing” a gunshot that dispatches a police squad to the scene of the crime. Why make employees’ risk their lives by pushing a panic button as soon as the robbery starts if the camera system can do that automatically – with Xeoma’s Sound Events Detector?
Home: not only is “Sound Events Detector” a perfect anti-break-in solution (especially when combined with Xeoma’s Motion Detector, Loitering Detector or Face Recognition) – it is also a new-gen baby monitor & nanny cam! It can hear baby cry and send you notifications and video extras of what happened prior to that! Besides, as it’s an all-in-one bundle, you will also get recognition of screams – for instant detection if the nanny raises their voice at the child while you’re not looking.
All done on your machine, no need to connect to any cloud servers.
Works with real-time video streams. The more fps, the better!
One-time payment. You don’t need to pay per image.
No special equipment or a powerful standalone server is required. Recognition load is on average 20% more than that of a regular camera in the program. Use our calculator here.
No special cameras with embedded recognition detectors required. Recognition processes are conducted on Xeoma’s side. High quality video stream is enough. Point the camera as perpendicularly as possible towards the faces, and make sure that the face is in the camera’s field of view for at least a few seconds.
Recognition of all 5 types of sounds in one pack – at affordable price. Depending on the required conditions and the desired result, you can choose Xeoma Pro or Xeoma Standard licenses and purchase the additional module “Sound Events Detector”. Check prices for Xeoma licenses and Additional modules here
Simple interface, fully automated work, high-speed processing and analysis. Just connect the “Sound Events Detector” module into the chain, select the type of sounds the system needs to pay attention to, plus the acceptable accuracy level of recognition, and you’re good to go!
|Please note that this module is available only in Xeoma for 64-bit Windows, Linux or Mac OSX operating systems. It is not available for Linux/ARM. Contact us if you need it for an unsupported OS.|
|Xeoma’s ‘Sound Detector’ is another module working with sounds – it analyses the volume of the sound.|
Thanks to the flexible modular system in Xeoma software, “Sound Events Detector” can be used in conjunction with other Xeoma modules or external systems.
For example, for security purposes you can use a chain of modules “Sound Events Detector” + notification modules (pop up window, email and SMS sender and so on) to be instantly notified of the alarming event.
Also, with the “App Runner” or “HTTP Request Sender” modules, customized reactions can be triggered (launching a specific external tool or a user’s script, or sending a command to external automation systems).
Combination with the “Move to PTZ preset” module can help rotate an overview camera with PTZ functions to face the area where the looked-for sound was detected.
Use with “Motion Detector” / “Loitering Detector” and “Scheduler” will form a simple and effective anti-burglar system for home or business in off-hours. Add “Face Recognition” or even “FaceID” modules for a more refined access and statistics on visits.
Utility filter modules like “Image Rotate”, “Privacy Masking”, “Marking” will help you adjust the system to show the picture in a convenient way, and the “Problems Detector” will keep you informed on the system’s well-being.
Over 40 more filters are available in Xeoma, and more are coming! See list of modules here
The Sound Events Detector module can save reports in spreadsheet CSV files (tick the corresponding option in its settings for that).
CSV spreadsheets are used to form a powerful reports system. You can import CSV reports into programs like Excel or LibreOffice and build reports and graphs of any sophistication level there. Alternatively, CSV can be imported into mysql and used to form any kind of reports with the help of mysqladmin web interface – or just make a simple several-lines-of-code script for an automated reporting. See examples and guides in the Internet.
Try Xeoma’s module “Sound Events Detector” with a trial license. Request it in our contact form here!
|Do you need something else? We can develop it and add it into Xeoma as the paid development. See details here or contact us|
May, 22 2020
Additional modules in Xeoma
Buy Xeoma’s Additional modules
Face detector (emotions)
More about Face Recognition
Required hardware calculator
“HTTP Request Sender”
List of all modules in Xeoma User Manual
“Move to PTZ preset” module